Is ‘Good’ Good Enough?

Years ago I heard the trainers I learned from talk about the 95% rule.  At the time is seemed such a huge goal to meet that I thought, until I got better at training and seeing the detail they saw, ‘good’ approximations to the movement I wanted from my horses would be fine for me.  However, with the re-emergence of discussions on resurgence and extinction, it has become clear that ‘good’ is not good enough no matter where we are in our knowledge.

The reason for this is quite simple.  What you learn first you learn best and that is, in times of confusion or stress or frustration, what you will default (or regress) back to.  It never goes away.  That first ‘good enough’ approximation in to a behaviour that was ok at that time always re-emerges and its why we have days where we say “yesterday he could do this so easily, why is he not giving me his best today?!”.

What this means for us as trainers is that we need to change how we approach the way we look at the behaviour we train.  Previously I described training a behaviour as accepting everything from A to D in one behaviour, even though we were aiming for B. That over time we could filter out and only reinforce A to C, then B and C, and then only B.  The idea was that as the learners knowledge and skill improved they could nail B more and more easily.

However, each time we were offered A to C, not only did we reinforce B, but we equally reinforced A and C (which we did not want).  That means that we have equal chance of, when we ask for B, getting A, B or C.  When we finally are just getting B, A, C and D have been reinforced so much that the difference between the number of reinforcements between those and B is so small that there is a high chance that A, C or D could re-emerge (resurge) right along with B.

When thin slices are built up a whole image appears.  The thinner the slices the more perfect the final image will be.

When thin slices are built up a whole image appears. The thinner the slices the more perfect the final image will be.

What we need to do instead is look only at B and think slice B.  This is really what MICROSHAPING is all about.  We take the shape we want and we train approximations of only that behaviour and teach a small piece of it at a time.

If we think about lifting a horses hind foot for cleaning.  If we were to take the previous approach then what we might do is go straight in and get the foot off the ground the horse may be off balance and they have to move their leg around a lot and  try to ‘take it back’.  That results in the handler hanging on waiting for the horse to stop so that we can reinforce.  But what we reinforced was foot lift AND a lot of other behaviour that involved potentially dangerous leg flailing.  All that ‘unwanted’ part of foot lifting is always there because it was reinforced.

The microshaping approach would be to think about what the very first thing the horse has to do to lift that foot off the ground; take the weight off that hip on to the other hip (not forwards).  The first microshape in to the whole behaviour is asking the horse if they can move their weight on to the opposite hip.  Then as we ask for more and more of that the behaviour we really want begins to emerge more clearly.  The heel of the foot begins to lift off the ground and eventually the toe lifts as well.  We have no unwanted variations in there because we set things up right from the very smallest approximation of the whole behaviour….starting with how the horse has to re-balance.

Regression is also why adults go back to old reaction patterns when parents behave towards them in a certain way, or say certain things. Previously reinforced behaviours resurge in the presence of the right set of circumstances.

Jesus Rozalez Ruiz also talks about this and he used a much more base example. If men, often when having a few beers, ever practice relieving themselves in the great outdoors then should they have the misfortune of having dementia later in life, will regress back to previously reinforced behaviours and may be found inappropriately relieving themselves in the garden!

As Alexandra Kurland says; don’t let your horse practice something you don’t want them to get good at.

Putting that in context with the new info we have on extinction and regression; don’t reinforce behaviours you don’t want to reappear later on in your training. Don’t consciously reinforce a behaviour that you will have to extinguish because the extinction process can be stressful and frustrating, therefore damaging to a relationship.

Even when a behaviour is extinguished there is a phenomenon called an ‘extinction burst’.  If something has been reinforced, given the right set of circumstances it will re-emerge even if you have not seen the behaviour for 10, 15 or more years.  We can never unlearn
what we learned (with the exception of as a result of brain trauma).

To enable microshaping to be successful training plans are very important.  Also understanding your environment and how to use it/manipulate it to set your horse up for microshaping success is vital.

Happy microshaping 🙂



Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s