The 300 Peck Approach to Building Duration
This idea is based upon an experiment in variable reinforcement schedules. A scientist was looking at how to build long duration variable schedules. She taught her pigeons to peck a bar, and by gradually extending the number of pecks they had to give her, she got her pigeons on a variable ratio schedule of 300. That means that on average they had to peck the bar 300 times before they were reinforced. That sounds like a tremendous amount of behaviour, and it is, but it is in fact what we want from our dogs. Think about dogs doing long stretches of heelwork in test C obedience or a dog carrying out a 10 minute downstay as in working trials or a dog that is required to walk on a loose lead for 20 minutes or more. All of these behaviours can be trained using this 300 peck approach.
As we all know, the secret in building duration (whether it is time or number of steps) is to slowly increase the time (or distance). This approach uses increments of 1 and builds them methodically up to 300, where by the dog has learned to carry out the behaviour for in effect infinite duration.
An integral part of this method is that the trainer counts out loud to the dog in a rhythmic metronomic cadence. Counting out loud is very effective for several reasons.
Plateaus are steps that you just can't seem to get past. This could be the dog offering you other behaviours to try and get the click and treat or the dog moving, barking etc, and breaking the original behaviour. It is important to know when these plateaus are happening – i.e. it takes several attempts to get past the count of 6, or 10 or 30. Plateaus can come at any time. Whenever the behaviour we are rewarding is interrupted and the criterion defined by us is no longer met, the count goes back to 1 and is built back up again.
This approach can be used to effectively build solid stays as well as exercises such as ‘watch me’, hand touch, loose lead walking, competition heelwork and so on. I have used it with great effect on dogs that are confirmed pullers and they have responded rapidly.
If we look at teaching loose lead walking, our only criterion is that the lead must be slack when we reach the count. It doesn’t matter what else the dog is doing when the count is reached (sniffing the floor, looking at you, wagging tail etc.) the only important thing is that the lead is slack. If the lead isn’t slack when the count is reached, then the count is reset to 1.
So the process looks something like this:
1 - C/T (C/T = click and treat for more information read this)
1, 2 - C/T
1, 2, 3 - C/T
1, 2, 3, 4 - C/T
1, 2, 3, 4, 5 - C/T and so on until you reach 300.
In fact, you will probably get 3 or 4 steps before the dog tightens the lead and you reset the count. You may have to reset the count several times before you progress past a count of 10. I usually begin with the food reward in my hand (but not at the dog’s nose height) so that the dog is aware that it is there and available for earning. When the dog keeps the lead loose for the required count, then the food is dropped just behind my legs. As the count increase, the food is delivered from a treat bag and isn’t kept in my hand.
You may walk 85 steps or more in order to achieve that count of 10 in the beginning, but as the dog grasps the idea; it will begin offering the behaviour more frequently and readily. Don’t be surprised if the dog begins to experiment and you get barking, sideways movement, the dog watching you or any other behaviour. Ignore them all and only work on the criterion that the lead must be slack. Rudi started dropping his head and tail, which had the added effect of being very calming. When he is more consistent in more areas, this criterion will be paired with the loose lead one. However it has already made him much calmer when he comes out of the van and he can now walk on a loose lead immediately, rather than me having to wait until he calmed down.
I have found this to be the only method that has worked for Rudi, my rather over excitable GSD that just didn’t grasp the concept of loose lead walking. He’d dash to the end of a 6ft lead and be back at my side before I could blink. Its taken a double connection lead (taken from TTouch) and the 300 peck approach for him to begin to grasp the concept. He’s excellent in familiar places, but we have to go drop the criteria still when we get to a new area as he finds it all so exciting. This method has also been very successful in teaching Rudi to stay.
As the count increases past 100, you’ll find that you can increase the increments a little faster, perhaps by a count of 10 instead of a count of 1. There will be some ‘sticky’ spots. One trouble spot seems to be around the 60-70 mark. Its almost as thought he animal is saying ‘yes I’ve got the idea now, can we do something else instead?’ work your way through it and keep counting.
Stays are built exactly the same way as the loose lead walking. Count aloud, it helps you to keep count and really does stop you rushing through and builds a solid stay. I start off standing close to the dog, and as the count increase past 60, then I will also add a little movement on my part; a slight sway from side to side, then a step to each side, gradually building to moving around the dog in two semi-circles. As the dog's stay increase, then I will begin to add a little distance, but I will also decrease duration to begin with; so I may drop down to a 20 second stay and walk 10 paces away, and walk straight back (roughly 20 seconds) and then build this up. Do try not to work on both distance and duration at the same time; only one criterion at a time.
Give this method a go and see how you get on. See how many different things you can build using the 300 peck method. Remember to be consistent…every time you take the dog out on a lead, then it must walk on a loose lead, you will soon begin to reap the benefits even if it seems like an uphill battle to begin with.
A pdf version of this page is available here