For an in-depth look at 300 Peck the way Sue Ailsby does it, click here.
Recently on the Training Levels list there was some discussion about variations on the 300 Peck Method, with one subscriber asking " Lets say the dog is doing well, walking 45 steps consistently or so (picking a random number here). Is there ever a point at which you dont start with rewarding 1,2,3? Would you then begin to start higher?"
Sue Ailsby replied "Sure. If you're up to 45, start back at 30. If you do that twice and
the dog breaks at 45 BOTH times, go back to 20 and start from there."
I suggested a variation which I often use, and that is to reset the count to 10% or 20% of the maximum. e.g, if you're up to 50 steps and the dog fails, restart at 10 steps. If you're up to 100 steps and the dog fails, restart at 20 steps.
You can also increase your criteria in increments of 2, e.g
1,2 click and treat
1,2,3,4 click and treat
1,2,3,4,5,6 click and treat
However, the most important thing is to maintain a high enough rate of reinforcement. When you use increments, or reset your counts to relatively high numbers, the rate of reinforcement drops off dramatically.
It could be argued that the rate of reinforcement when using the 300 Peck Method is too high for most long duration or long distance applications, and this is probably true, but you should see steady improvement. If you stop seeing steady improvement and you have changed something about the way you use 300 Peck, then your rate of reinforcement has probably dropped off.
I had this problem teaching my dogs to stay on mats together. I was using increments to raise my criteria, and ended up with highly variable points of failure for one of my dogs. She would break her stay at 15 seconds, then 24 seconds, then 30 seconds, then 12 seconds, then 38 seconds, then 7 seconds.
I wondered if this was a "feature" of 300 Peck, given that the criteria would reset to a low number on every failure? But this had not been the case when I was teaching loose leash walking or Go to Target using 300 Peck, where I saw steady improvement and only occasional failure at relatively low criteria.
In the end I decreased my use of increments to increase the rate of reinforcement and we were able to steadily increase our Stays on Mat.
But then the whole idea of 300 Peck is that you don't have to think about what your criteria should be. So if this feature of 300 Peck appeals to you, don't mess with it. Reset your counts to 1 on each failure, and increase your counts by 1 on each success.