"The Clicker Method"
Ask anybody training and competing in the sport of dog agility what is the most difficult obstacle to teach their dog, invariably they will answer that it is teaching the weaving poles, which can take many months, how true this is, few dogs really ever get to fully understand the weaving poles, mistakes are common, performance for many is a lottery.
Could this be a direct result of the way we train them?, or because of our perception of what weaving poles are? We try many different methods to get our dogs to do weaving poles, staggered poles forming a chute, V-shaped weaving poles to focus the dog taking a central path, all using multiple repetitions to hopefully shape with time the weaving behaviour we want. The most common result of all these training methods is a dog that "does the poles", but has not learned to "weave". With the science of Operant Conditioning and Positive Reinforcement being used more to-day in dog training, thanks to the popularity of books like "Don’t shoot the dog" by Karen Pryor (1984). With the use of positive reinforcement we can shape our dog’s behaviours and then back-chain them into a complex behavioural chains (DSTD p.96.), these same principles can also be applied to training our dog to weave. Using our dog’s own operant behaviours and shaping the weaving behaviour by reinforcing with the timely application of a conditioned reinforcer called "the clicker". This method is quicker, more efficient and makes full use of the dog’s own natural speed and body rhythms.CONDITIONED REINFORCER
The clicker is a precise and consistent "Yes" signal for our dog, the "click" lets the dog know at that instant in time his own behaviour has been rewarding. It is important to understand that the clicker gives the dog information. The clicker also acts as a bridging signal, bridging the behaviour with the reward, while the click can be inherently rewarding to the dog, it is best not to weaken the power of the clicker, reward your dog with a treat every time you click. What happens after the click is not important as the click ends the behaviour.
First condition the clicker as an conditioned reinforcer, by giving a "click" followed a second later with a "treat", use something really special and tasty for the treat, chicken, beef, hotdog, salami, etc., (dried dog food or kibble, is not very motivating). Don’t ask your dog to do anything, just click and treat, after a short while, if the treats are really good, you should see your dog alert when you click, you now have your conditioned reinforcer.TRAINING
We must first break down the weaving poles into its basic unit, teach our dog to weave this basic unit, then back-chain this behaviour into multiple weaving units, resulting in our dog "weaving" through the weaving poles.
The basic unit of weaving is two poles, so to start our dog we have to shape the behaviour of weaving through two poles. The advantages of shaping this basic unit of two poles are:
We can train the entry/exit behaviours simultaneously. All behaviours are homogenious for easy backchaining. We get the basic weave behaviour first, the cue goes on a known weaving behaviour.
Eliminates superstitious cues.
The finished behaviour of weaving is independant of the number of poles.With your clicker and tasty treats at the ready, your dog off lead, you are ready to start.
STEP 1.
Set up two poles so that our dog can easily run between them, click and treat when your dog is between the poles, repeat a number of times: ( C/T = "click" and then treat )
o
o
dog --------> dog ------->
o
o
C/T C/T
When you have your dog looking for the poles to run between gradually reduce the distance between the poles to the normal weaving pole spacing, continue with C/T every time your dog is between the poles. STEP 2.
When your dog is happily earning his C/T running between the poles start the shaping process the by gradually rotating the two weave poles anti -clockwise, make sure your dog is happily looking for the correct entry to the poles before progressing.
-----> ------>
o o / o / o
dog --------> dog ------ o dog ---
o
C/T C/T C/T
When you raise the criteria for C/T, do it gradually, shape a very small step at a time. This can cause a fluctuation in offered behaviours, make sure you discriminate and only reward the correct weave behaviour;
--> ^ <----
o / o o | o o \ o
dog ------ dog ---- dog
C/T NO C/T NO C/T
(correct) (incorrect) (incorrect)
STEP 3.When your dog is fluent with this weave behaviour, and the poles are a "cue" to weave, the nest step is to condition your verbal cue "weave" as a conditioned reinforcer by giving the cue followed by C/T a number of times. Now introduce the verbal cue to the exercise:
----->
o / o
dog ----------
"weave" C/T
Repeat STEP 3 with different approach and exit angles.
dog ---------- ----
"A" \ o / o \ "B"
---- ---->
"weave" "Click" Treat
When we have the dog fluent with this we can now start to backchain these basic behaviours into weaving the poles. It helps to always start at point "A" and finish at point "B" so that a reinforcement history is built up that will assist the initial backchaining.STEP 4.
Set up two units of poles, place the new unit in front of the one used in Step 3. The first pair of poles act as a stimulus or 'cue' for the first weave behaviour. As your dog passes the second pole give the cue "weave", C/T as the dog passes the fourth pole. The cue "weave" then becomes the conditioned reinforcer for the first weave behaviour (poles 1 & 2), and the 'cue' for the second weave behaviour (poles 3 & 4), thus linking the two weaving behaviours together.
----- -----
o / o \ o / o \
dog ---------- ----- ----->
"weave" C/T
When you have a solid behaviour of your dog weaving move the cue "weave" to the first (new) weave unit of poles in the chain, and repeat:
----- -----
o / o \ o / o \
dog ---------- ----- ----->
"weave" C/T
STEP 5.Repeat the above (step 4) each time adding a new weaving unit till you have back-chained a full set of poles. Each time the verbal 'cue' is the reinforcer for the first unit of weave behaviour, and the "cue" for the following chained weave units. As with the earlier steps, make sure your dog is happily weaving the chained weave units before adding another weave unit into the chain.
---- ---- ---- ---- ---- -->
o / o \ o / o \ o / o \ o / o \ o / o \ o / o
dog ---- ---- ---- ---- ---- ----
"weave" C/T
With this method it is important to allow the dog to learn to weave without the handler distracting him, no corrections, no superstitious body language during this learning process, no extra verbal commands, let the clicker do the talking. Use random jackpots (one click, but lots of treats) to reinforce really good weaving behaviour.Up to this point you have trained your dog on a continuous schedule of C/T when back-chaining the weave behaviour, now your dog is performing the full weave behaviour its time to vary the reinforcement. Start with only C/T'ing every second correct behaviour of weaving, use a treat with a verbal conditioned reinforcer "good" (conditioned just like the clicker), then as you progress, every third behaviour, this will allow you to fade the Clicker, but always reward good behaviour, use treats, verbal conditioned reinforcers, and any rewarding activity, ball, or more agility obstacles to always reward good weaving behaviour.
When you have your dog doing the weaving behaviour for the full set of weaving poles, you can start to train in a course context by putting an obstacle (usually a jump) before and after the weave poles.
Have the jumps end on to the weaving poles at first, then train angled approaches, observe normal spacings between obstacles.
jump jump
| |
dog --- | ---------- o o o o o o o o o o o o ------- | -->
| |
"over" "weave" "over" C/T
If your dog makes a mistake do not try to "correct" it, verbally nor physically, this may reinforce and "mark" the incorrect behaviour, introducing apprehension in your dog at the point where the mistake happened, It is better to stop your progress, and not attempt the next obstacle, repeat the exercise, temporarily introducing the cue "weave" to encourage your dog just before the point of error to help get past it . It is useful to use the Clicker, on occasion during training to sharpen up behaviours that may have 'drifted' over time. Only click for behaviour that meets your criteria.
Hope you find this method of teaching your dog to weave easy, quick, and fun too!
Happy training,
Robert Loftus
References:
Karen Pryor, (1984)
Don’t Shoot the Dog, ISBN 0-553-25388-3. *New Edition escaped August 1999.
Training Weave Poles: The Clicker Method
by Robert Loftus was originally published in
Clean Run (USA), February 1998 and has been reprinted with the
permission of:
Clean Run Productions, LLC, 35 Walnut Street Turners Falls, MA 01376, USA
Phone: 413-863-9243 or 1-800-311-6503; 24-Hour Fax: 413-863-8303
Email: info@cleanrun.com or visit: http://www.cleanrun.com/
Clean Run Productions, LLC publishes the Clean Run (USA)
magazine along with an
assortment of agility-related books. See their web site for more information.
V 1.04
İRLoftus1997