09/20/2026
One of the more common “perceptions” people have of Positive Reinforcement is a horse being clicked and fed every few seconds for the same behaviour over and over.
To be fair, some people do train that way, but that’s not what shaping normally looks like.
Regardless of what method you train with, we often break behaviours down into small steps until we get to the “finished” or desired behaviour. This is what we call an “approximation”, and that means we gradually raise the criteria and shape through approximations to get to our finished behaviour.
Sometimes that progress can be tiny. Sometimes you repeat a step a few times because the horse needs more confidence or isn’t quite “there yet”. Sometimes we even go back and lower our criteria, but overall, we want to see the behaviour keep DEVELOPING.
While there may be instances where I reinforce something as small as a weight shift, a lean, or even slight engagement, I am not doing that for every single step. That level of approximation is usually reserved for specific situations.
Most horses are able to take bigger or smaller steps toward the goal depending on the behaviour, their history and how well they understand the training. ESPECIALLY horses with a solid R+ history who already understand the “game”.
When we are months or years into training, our horses understand how shaping works. They understand duration, distance, building criteria, and that we are working toward something.
If I have clicked several times and there is very little change between each repetition, and the behaviour is not progressing in any real way, I am going to stop and look at why.
Progress does not always mean the behaviour looks dramatically different. Maybe I am building duration. Maybe I am getting more distance. Maybe the response is becoming more precise, more confident, more fluent, or happening around a harder distraction. But SOMETHING should be developing.
And the approximations will not look the same for every horse. They can and should change depending on the individual horse, their history, confidence, experience, and what you are training. A good trainer needs to be flexible enough to adjust the size of those steps as they go.
Let’s use training haltering as an example because any horse person will know this behaviour.
When we are teaching a horse to halter, there are a series of things involved to get from never seeing a halter to wearing a halter. A horse that is feral or unhandled will likely take more steps than a horse that’s been handled since day one.
Before we even get to the halter, we need some behaviours to already be solid. They need to be able to stand comfortably with me in their space, they need to understand what the click means, they need to understand expectations around food and cannot be mugging or constantly throwing behaviours, and depending on the anxiety level in the horse, I will ask for a start button.
THEN we can break down the haltering itself:
• touch the halter
• bring your nose toward it
• put your nose in
• tolerate the halter being moved, making noise and being in your peripherals
• tolerate it moving around your ears
• allow me to tie it
• add in the lead rope
I’m not going to keep reinforcing “touch the halter” forever. That behaviour is just ONE approximation out of many on the way to the finished behaviour of haltering.
The same applies to all other areas. When a horse understands something, we should be looking at how to build that behaviour further through things like duration, distance, precision, movement or chaining in other behaviours.
R+ can absolutely become stagnant if the trainer never advances criteria. But once again, that’s a TRAINING issue, not a requirement of Positive Reinforcement.
