Improving Gaits Through Clicker Training
Someone asked the following question in response to my post on clicker training and smooth trots:
"I'm still confused as to the 'how' of clicking then giving reward to horse for correct trot (or any other correct move/motion), without interrupting the flow of that motion. If you stop the horse to reward, I envision a horse that stops whenever he hears a click?"
This is something that trips people up until they become familiar with the learning process the clicker represents. Yes, the horse does stop when he hears the click, and that's exactly what we want. The clicker builds on the power of variable reinforcement schedules.
We probably all were originally taught that we had to keep the horse going no matter what. We weren't supposed to let the horse break gait. But we were also taught how important transitions are, and that's what the clicker gives us, opportunities to practice literally thousands of transitions. That's what you need to create a safe horse, and it is also what you need to produce upper level performance.
So how does the clicker work? Let's go back to the order in which collection, (or better yet, connection) develops. The first thing you get is a collected/connected stop. Then you get a collected/connected start, then you get a collected/connected middle, and finally you make the collected/connected middle longer by putting more strides between the start and the stop.
If this sounds backwards, or confusing think about all the discussions on this list about disengaging the hindquarters, and one rein stops. When does this training begin? The very first day you ever ride the horse, and even before that on the ground.
On the ground one of the first things I teach a horse is to back up. I use the clicker in this early stage. Of course, I can teach backing without the clicker, but if I do, I'm missing an opportunity to show the horse how the clicker works. If I don't include it in these simple lessons, he won't know how to use the information it provides later when the work becomes more complex.
When I teach a horse to back using the clicker, I'm also employing all the principles I'll need to use the clicker under saddle to improve the trot or the canter. I'm going to work in a series of stair steps where I keep asking for more from the horse. I'll begin by triggering the back in the same way I would if I weren't using the "yes" answer signal of the clicker.
If you're familiar with Parelli's Seven Games, you know how to ask the horse to step back from pressure on the bridge of his nose. Or perhaps you're using Lyons' "college level leading" lesson, and you're asking the horse to step back from the tap of the whip. Both are using a negative reinforcer to trigger the response. A negative reinforcer is an unpleasant or painful stimulus which the horse can avoid by changing it's behavior. The key here is that as soon as the horse even so much as shifts it's weight back, the pressure from your hand, whip, or lead is removed. The horse learns that it can get you to stop simply by stepping back.
I add the clicker to this. As the horse steps back, I remove the pressure, plus I click and give the horse a treat. Now the horse isn't just moving away from discomfort. He's working for something he wants. The clicker accelerates learning and it creates enthusiastic, can-do horses. It changes the nature of the negative reinforcer. Pressure isn't a threat. Instead it becomes information the horse uses to sort out faster what will earn it a treat. (For Lael with her mustangs this is an important aspect of the training. The clicker is a teaching tool that helps build trust and takes the fear and intimidation out oftraining.)
The next step in this is to ask for more than just a shift in weight. Now the horse has to take a full step back before it gets a click and a treat. I can quickly up this from one or two steps to three or four, and then a whole line of steps. I'll use the initial tap of the whip or flick of the lead as a "green light" cue that says you may now back up. As the horse begins to back, I'll turn off my cue, but the horse is to continue backing until I either ask for something else, or I click. If he stops prematurely, I'll ask again with another tap of the whip. I don't want to have to tap him continuously. This would be like having to ride with my legs constantly banging on the horse's sides to keep him moving.
So I started on the ground with a horse that didn't know how to back. Maybe it was either braced and glued to ground, or pushing through me, and in just a few minutes using the clicker it is now freely and willingly backing the length of my arena. That's the beginning of a collected stop. I've got a horse who will on my signal shift his weightback.
The next step is to ask the horse to go forward. Instead of pulling the horse forward from the front end, I'll get the hindquarters to move. Again, I'll use pressure to trigger that response, and I'll reinforce it with the clicker. If I'm using Lyons' college level leading, I'll tap the horse's hip. That becomes my cue spot to ask for forward movement. Just as with the backing, I'll be satisfied at first with a simple shift in weight, and then I'll gradually ask for more. What I am learning is how to use a variable reinforcement schedule. This means that the horse is never sure exactly which step is going to get reinforced so it keeps on offering more behavior in the hopes of earning a treat. Slot machines work on this principle. People pump millions of dollars worth of quarters into them every year, so there must be something to all this. (I wrote a post to the list recently on variable reinforcement schedules called "pecking pigeons and riding on contact". If you would like a copy, email me privately.)
With these principles and basic exercises I have everything I need to build good gaits under saddle. If I can get a horse to shift its weight back, and to shift it forward, I can connect its feet to the reins, and teach it to soften its topline.
I'll begin by simply letting the horse do what it wants to under saddle. If that's to stand planted to the ground, that's great, and if it's to rush off in a nervous, high headed trot that's also great. I'm going to do the same thing with both horses. Before the first one falls asleep, or the second one gets out of control, I'm going to pick up the rein and ask the horse to move its hips away from the direction of the bend. I may do this in stages. I may ask for the jaw first, but I'm headed for the hips. That's going to get the planted horse to move, and the nervous horse to stop. I'm going to repeat this request many times until yielding the hips becomes a smooth, easy response. In the process I'll find that the jaw, neck, and shoulders will have softened as well. Instead of riding a braced, rigid, pulls-like-a-freight-train horse, I'll find that I'm sitting on a light, soft, relaxed horse.
This is the beginning step of Lyons' "hip, shoulder, shoulder" exercise which produces a collected stop, and teaches the horse how to lift itself up into the equilibrium of an upper level performance horse. Each time I bring the horse to a stop, whether it's through a one rein stop, or the clicker, that gives me an opportunity to work on my "go forward" cue.
That's so important, because every time I pick up the rein and ask the horse to soften and yield some part of his body to me, I'm bleeding off energy from the horse. That is, after all, how you get stops in the first place. If you keep asking for changes of direction, pretty soon your horse will be slowing down and standing still. So, if you don't know how to put energy back into a horse, you're going to be in for some very quiet rides, as in your high energy horse is now the one standing with his feet planted in cement. So, teaching the "go forward" goes hand in hand with teaching the horse a one rein stop.
So here's the process. My horse is walking along in an easy going walk. I pick up the rein and ask the horse to yield his hip. At first it will help us both to learn this if I look back at the inside hip. I want to see the point of the hip swing away from the direction of the bend. That means if I'm on the left rein, I want to see the hip swing to the right. The instant that I see even a little movement, I'll click my horse, plus I'll release the rein. The clicker is a bridging signal. It marks exactly what I want to reinforce. The horse associates whatever it was doing at the exact moment when the click sounds with the treat it is about to receive. (I make a clicking sound with my tongue instead of using a mechanical clicker so my hands arefree.)
The horse is going to want to reproduce that behavior again. So the next time I pick up the rein, it will be trying to figure out what it just did so it can reproduce it again. Just as I built backing on the ground from a simple weight shift to multiple steps, I can build this response into the larger step that creates a halt. Click! My horse is going to get a jackpot for stopping.
I've already been practicing my "go forward" cue because each time my horse stopped after I clicked, I had a chance to work on this very important element of his training. I keep repeating this cycle. I use my "go forward" cue to initiate movement. I let my horse go a short distance, then I pick up the rein and ask for a "hip, shoulder, shoulder" one rein stop into a step or two of soft reinbacks. Click! he gets a treat. As this cycle becomes consistent, I'll add a new element to the sequence. Instead of clicking after the soft rein back, I'll shift my seat to ask for forward movement. My horse will probably ignore me at first. He's so focused on the pattern, that he doesn't notice the change in the cue. Backing was the right answer before, so it must be the right answer now. People do exactly the same thing when they play the trainer's game. It's just part of the learning process. I just let the horse keep backing until he notices this isn't working. He hasn't gotten clicked, and my seat is still asking for something. The instant that he changes direction, Click! he gets a treat.
I'll just keep working on this sequence until the instant I change the intent of my seat, my horse responds. (Let me anticipate one of the most frequently asked questions here. People will ask what are you doing with your seat. Even if I tried to answer that here, it wouldn't mean much. The real answer comes from riding through the steps of this process. That's the real brilliance of all these chunked down single rein riding exercises. These steps are doing more than just training the horse. They train the rider as well. If you don't know the feel you're looking for, go ride a thousand single rein stops, or better yet, five thousand. That's all it takes to answer that question - the actual doing of it.)
So now I have a horse that stops on a very light rein into a soft, collected shift of weight back, rocks back onto it's hocks, and then shifts its momentum forward when I ask it to walk on. That's the beginning of a collected start. The next step is to ask it to continue on in whatever gait I've asked for. This is the beginning of a collected middle. I don't want the horse to just fall apart and get all strung out in the walk. I want him to carry all that beautiful engagement and connectedness forward into the walk. As he gives me that extra step or two that retains the elements that the stop and start have created, click, he'll be reinforced. Tie this into the pigeon post on variable reinforcement schedules, and you'll see how you can develop outstanding gaits. And yes, your horse will be stopping, but he'll also be learning what elements earn reinforcements. Just as you did with every other stage of his training, you'll increase your standards until he's giving you a beautiful collected start followed by a collected middle of extended duration, and finishing with a collected stop. If at any point in this cycle, you want to highlight and reinforce his performance, you can do so easily with the clicker.
This process is best learned first in the walk (see clicker training and smooth trots), but all the principles apply to the trot and canter. The beauty of this process is you're beginning with the building blocks that make horses safe to ride. The single rein stop keeps horses from bolting, bucking, and spooking. By perfecting that exercise you end up developing smooth, spectacular gaits. When you add in the clicker you create a learning process that is fun for the horse, and very rewardingfor you.
Judy asked a short question, to which she got a long answer. My apologies for the length of this post, but the only way to answer this question was to go back and look at the preceding layers in the training. If you don't understand the building blocks, you won't understand the answer. The easiest place to learn about weight redistribution is on the ground in the beginning leading exercises. If all of this seems confusing, begin there and build your layers systematically. They'll take you as far into this process as the goals for your horse need you to go.
The printed information contained in this fact sheet is kindly provided by JudyRyder Duffy http://clickryder.8m.com/, and Alexander Kurland; author of "CLICKER TRAINING FOR YOUR HORSE" GETTING STARTED: CLICKER TRAINING FOR HORSES. You can purchase Alex's books by printing off the order form at http://www.crisny.org/users/kurlanda