To be most effective in our shaping, we want to strive for continuous incremental progress. When we are doing it at our very best, we increment every rep, but in such small increments that the dog is always winning. Easier said than done!
We want those increments to be so small that the dog always has a good chance of being correct. And we can be reasonably sure that that next version of the behavior is going to occur.
We don’t want to plateau, because the longer we spend at any one stage of training, the more ingrained that version of the behavior becomes and the harder it is to make those small changes that will result in progress, and increases the risk of confusion and frustration down the road. Of course, you know the danger of jumping up the steps too quickly… increased error rates, and the resulting confusion, frustration, and loss of motivation. So we’ll shoot for that happy medium. Which of course is the hardest state to reach, amiright?