Forward Chaining, Back Chaining, and Total Task Training
If you've ever seen some of the more complex tricks out there (like painting or fetching from a purse), you might wonder how in the heck did they train those?! Well, the best answer is that after splitting each trick into tiny steps and having the rat master them to perfection, we combine the steps back together. This might sound a little confusing, but think of it like a kid learning to write an essay. First you teach the kid to read. Then you teach the kid to write. Then you teach them grammar. Finally, when it comes time to write the essay you combine all 3 "behaviors" to get the final product!
Its the same with animals in that breaking a trick up into small bits and training them separately, then adding them back together is the best way to train any complex trick. On this page I will go over 3 different methods of training multi-step tricks, and then give you my take on which is the best way to get a reliable behavior as fast as possible.
Its the same with animals in that breaking a trick up into small bits and training them separately, then adding them back together is the best way to train any complex trick. On this page I will go over 3 different methods of training multi-step tricks, and then give you my take on which is the best way to get a reliable behavior as fast as possible.
Total Task Training (TTT):
In this training method you want the rat to complete the final behavior right from the start. For example, when training spin many rats can be lead in a full circle and then rewarded. This behavior (a full circle) is what the end result will look like, and works well because spin is a simple trick with really only one step. For more complex tricks like agility courses though, TTT is much less useful. You can technically train agility with TTT, but since this means that you aren't training each obstacle first and that the rat is solely reliant on the reward at the end of the course, training will be much slower, more frustrating, and your rat is more likely to decide that since they only get rewarded by the last obstacle, they will ignore everything else and walk straight towards the end.
In this training method you want the rat to complete the final behavior right from the start. For example, when training spin many rats can be lead in a full circle and then rewarded. This behavior (a full circle) is what the end result will look like, and works well because spin is a simple trick with really only one step. For more complex tricks like agility courses though, TTT is much less useful. You can technically train agility with TTT, but since this means that you aren't training each obstacle first and that the rat is solely reliant on the reward at the end of the course, training will be much slower, more frustrating, and your rat is more likely to decide that since they only get rewarded by the last obstacle, they will ignore everything else and walk straight towards the end.
Forward-Chaining (F-C):
In this training method you teach the rat every step/obstacle separately first. Once they can do each step/obstacle without any assistance, you begin to add these steps together by having your rat navigate the first obstacle/step, then rewarding them. After that, you have them perform the first obstacle/step, give them a treat, cue them to perform the 2nd obstacle/step, then reward again. After a few repetitions, you'll fade out the in-between treat and only reward after the 2nd obstacle/step has been completed. You'll continue to add in obstacle/steps until your rat is doing the whole trick/course for one treat.
Now forward-chaining can work really well for some tricks, but since it has your animal working towards something they've practiced less, it can be more frustrating and confusing on both the trainer and animal's parts.
In this training method you teach the rat every step/obstacle separately first. Once they can do each step/obstacle without any assistance, you begin to add these steps together by having your rat navigate the first obstacle/step, then rewarding them. After that, you have them perform the first obstacle/step, give them a treat, cue them to perform the 2nd obstacle/step, then reward again. After a few repetitions, you'll fade out the in-between treat and only reward after the 2nd obstacle/step has been completed. You'll continue to add in obstacle/steps until your rat is doing the whole trick/course for one treat.
Now forward-chaining can work really well for some tricks, but since it has your animal working towards something they've practiced less, it can be more frustrating and confusing on both the trainer and animal's parts.
Back-Chaining (B-C):
Like in forward-chaining, in this training method you teach the rat every step/obstacle separately first. Once your rats can do each obstacle without help, then you'll move on to having them do the last obstacle for a treat. Next, you'll have them do the next-to-last obstacle for a reward, then the last obstacle. After a few tries, you'll fade out the in-between reward and only reward them for doing both the next-to last obstacle/step and the last obstacle/step. And you'll just progress backwards from there, until your rat is doing the entire trick/obstacle course.
Back-chaining is probably the best (and most efficient!) way to teach any sort of multi-step trick, simply because you have the animal working towards what they've practiced the most. Rats (and any animal really!) are like humans in that they like completing easy and well-known tasks. So by having them work towards something that they know well and find easy, that actually adds a second reinforce for them to continue doing the trick!
Like in forward-chaining, in this training method you teach the rat every step/obstacle separately first. Once your rats can do each obstacle without help, then you'll move on to having them do the last obstacle for a treat. Next, you'll have them do the next-to-last obstacle for a reward, then the last obstacle. After a few tries, you'll fade out the in-between reward and only reward them for doing both the next-to last obstacle/step and the last obstacle/step. And you'll just progress backwards from there, until your rat is doing the entire trick/obstacle course.
Back-chaining is probably the best (and most efficient!) way to teach any sort of multi-step trick, simply because you have the animal working towards what they've practiced the most. Rats (and any animal really!) are like humans in that they like completing easy and well-known tasks. So by having them work towards something that they know well and find easy, that actually adds a second reinforce for them to continue doing the trick!
Trick-training wise, back-chaining is definitely the way to go when training complex tricks! |
Watch the video below to see a demonstration and comparison of TTT, F-C, and B-C!
Disclosure: Esther (and her site rattrix.weebly.com) is a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for sites to earn advertising fees by advertising and linking to amazon.com