Wednesday (Day Six), OK we started the day with the two sets of 2x2s 5 feet apart at 1-7 o'clock.
Sadly the day did not go well. Enya was missing the entries from the left again. She was bending round the first pole perfectly, but then skipping to the end and going between poles 3 and 4.
Initially I assumed this was cause by her speed. However when she slowed down she was still showing the same behaviour. Another indicator that there was more going on here, was her response to the "No" command (see "No as a command") when she did this. This should have caused Enya to try a new behaviour, instead she kept presenting the same behaviour. This suggests she thought the behaviour was correct.
I suspect in classic 2x2 training, it would be said that I had closed the gap between the two sets of 2x2s too quickly and I hadn't rewarded enough in-between the two sets as I closed the gap.
However I think there's an alternative explanation - I think Enya was offering exactly the behaviour I had trained her for.
Figure 1 above, shows the path I wanted Enya to take. Figure 2 shows what Enya was offering me. Figure 3 shows what I have been teaching Enya on Day One. If you had to describe the path in Figure 3 in words you could say she was to wrap the first pole, and then cut in before the next pole. Doing that would have produce the behaviour I wanted when repeated with another set of 2x2 as shown in Figure 1. However you could also describe path in Figure 3 as, wrap the first pole and then cut in before the last pole in the set. If you do that, then the path in Figure 2 makes perfect sense. She's taken what I've taught her on Day One with two poles and expanded it to four poles. That would explain why she was repeating it despite the use of "No".
So, whatever the explanation, we needed a cure. You're not supposed to make the challenge easier, but you do need to introduce a change to get the behaviour you desire. Roll on Day Seven.
Thursday (Day Seven). To start with I reduced the spacing on the 2x2 sets to 4ft, but re-opened the sets to nearer 2-8 o'clock. From the right hand side I was getting the behaviour I wanted. From the left I was still getting the skipping behaviour I'd seen the day before. Enya was actually deliberately widening her path to skip round pole 2 even when it was moved from the 1-7 o'clock to the 2-8 o'clock position. This further reinforced the idea that she felt this the behaviour was the correct on as she was actually avoiding cutting in-between poles 1 & 2, even when it was made easier. So as a temporary measure I moved the second set of 2x2s off the the reward line to the left. This meant that as Enya arced round she wouldn't see the second set unless she cut in-between poles 1 & 2. That had the desired effect, Enya started cutting in between 1 & 2 again, and then doing 3 & 4. Once you have the correct behaviour you can reward and reinforce it and I was able to move back to the normal configuration.
Because I was only getting the problematic behaviour from the left of the weaves I wondered if my body language could have contribute to the problem. The angling of the 2x2 sets on the reward line was not something I felt comfortable with. Your always fiddling with the two sets of poles to make sure the angles are the same, and the reward line is not consistent with the line you'd use for a normal set of weaves.
On Thursday night I suddenly realised there was a different way you could think of the layout of the 2x2 weaves. The actually geometry is identical, all that changes is the way you think of it.
Figure 1, below, shows the normal way of looking at 2x2 weave geometry. Figure 2 shows an alternative way of looking at it. The actual layout is identical - all I did was rotate the image in Figure 1 by 20 degrees or so. However it gave me a view of the weaves I was far, far more comfortable with. The baseline in orange is the same line as you'd normally see with a standard set of weaves. In this view it was immediately obvious to me that sets one and two are always parallel to each other.
It also makes adjusting the angle between the first and second sets of 2x2s far easier (at least for me). As the arrows show in Figure 3 all you do is move the second set towards or away from the base line. The only thing you need to be aware of, is that in this orientation the baseline stays still, but your reward line will move when you adjust the angles between the sets of 2x2s.
Tune again for the results of this change of mind set - did it help me or hinder? Was it a great idea or a disaster ?
No comments:
Post a Comment