Training Contacts

Of all the possible faults in agility, it seems that missed contacts are the most 
prevalent.  They are the hardest "issue" to retrain because what a dog learns 
first, he learns best.  To complicate matters, in the heat of competition, when 
the dog is cranked up, they tend to revert back to what they know best.

The first step in training the contact obstacles is for the trainer to have a 
clear idea of how the finished product should look like.  If there isn't a goal, 
then it's hard to achieve it.  It's very important to have a clear idea of how 
the dog should perform, so that his performance can be evaluated and rewarded 
appropriately.  Unclear goals lead to inconsistent training and rewards, and 
ultimately end up as inconsistent performances and missed contacts.

What kind of  'contacts' do you want your dog to have?  There are several 
types.  The most common type for novice trainers is the point and pray.  
This is not very reliable.  Then there is the running contact, the 'two 
on two off' contact and 'target' contacts.

The point and pray type of contact is an evolution of the dog's learning 
process and confidence building.  If you have no particular type of goal, 
you will usually end up with this type of contact.  And unfortunately, 
your prayers aren't answered very often, especially with an aussie.  Contact 
missing dogs aren't born that way.  They are trained that way by their 
unsuspecting trainer.  A trainer will teach the dog to go up and over the 
dog walk.  When the dog can perform this obstacle, they will say "There, 
my dog knows the dog walk."  The dog knows how to walk up and over the dog 
walk, but the issue of the contact has been totally ignored.  At first, 
the dog is successful in placing at least one paw in the contact because 
he is still unsure about what he is doing.  As he gains confidence and speed, 
he starts striding over the contact and missing it.  Even worse, the dog 
inadvertantly learns that jumping off the dog walk above the contact is 
good.  When the dog completes the dog walk, the owner praises him profusely.  
The dog associates being at the bottom of the contact (ie. on the ground) 
as a good thing.  So soon, the clever aussie puts two and two together 
and figures, "If I get to the bottom as fast as possible, I get cookies 
faster!"  The dog leaps off the dog walk to get to the bottom faster 
and another contact missing dog is born.

By the time the owner starts training their second agility dog, they have 
figured out they need a plan to address contact training before they begin 
training the contact obstacle.  Training the dog to go up and over the 
contact obstacles is the easy part.  Specifying exactly how they do it, 
takes a little thought and training.  But it is easily done when starting 
with a good plan.

One common 'controlled' contact performance is called 'two on two off'.  
This is the method I teach my dogs.  It allows for greater control.  
However, it can be slower than some of the other methods.  The 'goal' of 
this contact performance is to have the dog descend the down ramp of the 
contact obstacle and stop with his front paws on the ground and his rear 
paws on the contact obstacle.  Given the positioning of the paws, the rear 
paws will be in the contact zone.  The dog is trained to wait for a cue 
(usually verbal) to be released from the obstacle.  I have taught Hank to 
do this and after two years of competing, he has missed one contact.  I 
released him before he had a paw in the contact zone.  I was in contention 
for HIT and was trying to beat a border collie.  He was just doing what he 
was told.  

I like the two on two off contact performance.  One of the reasons is 
because it g ives me a chance to catch up and recollect myself if needed 
on the course.  Since the dog stops and waits for a cue to go, you can 
reorganize yourself if needed.  It also allows me to handle from a distance.  
I don't have to be right next to my dog to point and pray.  This is 
particularly helpful in gamblers.   Hank will go to the bottom of the contact 
zone and wait for me to tell him the next obstacle.  Sometimes, I will wait 
to release him until I see him visually lock in on the next obstacle.

Another reason I like the two on two off method is, when trained correctly, 
it teaches the dog to be aware of his foot position (see below).  The dog 
learns that his paws are what controls the dispensing of treats.  He is 
very careful where he steps.  And when trained this way from the beginning, 
he is very reluctant to leap off the contact obstacle for fun.

Another faster contact performance is called 'running contacts'.  I have 
seen varying success with this technique.  If you can get your dog to do 
it, it is usually faster than the more controlled methods.  However, it 
sometimes fails due to extra adrenaline causing the dog to leap or have 
a longer stride than in practice.  If you handle your dog where you will 
always be beside him as he is performing the contact obstacle, this method 
might be successful.  This method can be more reliable with a smaller dog 
who has a smaller stride than with a bigger dog.  

Running contacts are typically trained with hoops at the end of the contact 
obstacle.  The dog is trained to go through the hoop.  Since the hoop is 
about 2 feet tall, the dog can not jump off the obstacle.  If he passes 
through the hoop, then it forces the dog to run all the way down the 
obstacle, hitting the contact zone.  This method relies on muscle memory.  
However, in the excitement of competition, the dog's stride can change and 
his performance will be affected.

Contacts trained with targets are pretty popular as well.  I do not care 
for them though, although people do have success with them.  My main 
reason for not liking them is the risk of inadvertant training if the trainer 
is not saavy.  Targetting a contact means that a 'target' (usually a plexi-
glass square or a plastic lid) is placed on the ground in front of the contact 
obstacle.  The dog is trained to descend the contact obstacle and touch his 
nose to the 'target'.  A clicker is used to indicate that touching the target 
is the desired behavior and a treat is given.   Weaning the dog off touching 
the target and stopping at the bottom is tricky.  In addition, the dog is 
not aware of his foot placement.  The dog is focused on finding the target 
and misses the point of the contact zone. 

Some trainers will put food on the target and allow the dog to get the food 
when he descends the ramp.  As the dog eats the food they say "target!".   
They are attempting to associate being at the bottom of the contact obstalce 
with the word 'target'.   Many times I have seen a dog leap off the contact 
obstacle, pounce on the food and gobble it up.  A contact jumping dog is 
born and more importantly rewarded for it.  Then the trainer picks up 
the dog, places it's feet back on the contact obstacle and says "target!".  
The dog has no idea what the trainer is doing and only knows that it got 
a cookie when it leaped off the obstacle.

When you begin to train contact obstacles, you should always train the 
contact zone first.  Don't worry about teaching the dog to go up and 
over the contact obstacle.  That is the easy part.  First teach the dog 
the desired final performance and work backwards from there.  This is 
called back chaining.  Teach the dog the desired contact performance 
first and then teach them how to do the contact obstacle.  This way the 
dog will always be working towards what he knows best, since he learned 
the end (contact zone) first.

The first thing I teach a new dog is to lay down.  I teach the dog that 
this is the default behavior.  I never force a dog into a down.  I 
always train off lead and let the dog offer behaviors.  If the dog doesn't 
offer a down, I lure him into a down and give him a cookie. I will lure 
a maximum of three times and see if the dog will start offering that 
behavior.  If so, I give him a cookie for it.  I keep doing 'downs' until 
the dog offers it whenever I have a cookie.    This usually takes a 
few days of 10 minute sessions each day.

Next I will bring out the dog walk and begin with that.  I stand in front 
of the dog walk about 2 feet from it.  Since the dog has already been 
reinforced for lying down in front of me, they tend to try it again.  
My goal for this session is to get the dog to lie down on the end of the 
dog walk.  My ultimate goal is to get the dog to lie down with his front 
paws in the grass and rear paws on the dog walk.  If I have a young dog, 
then I will accept attempts that are close.  If I have an older dog 
that is pretty clever, then I will try to get close to the final performance.  
For the first session, I will reward any type of lying down on the board.  
The dog will be focused on me and will inadvertantly lie down on the 
end of the board.  I give the dog a cookie.  Then I will rotate my shoulders 
a little bit like I'm going to walk away and say "Ok!"  The dog usually 
leaves the board and follows my shoulders.  Then I stand back in front 
of the board to start again.  If I see the dog catching on pretty quick 
that lying on the end of the board is the desired performance, then I will 
make it a little more difficult and require that the dog lays with his 
front feet on the ground and rear feet on the board before I give him a cookie.

I only train this for a few minutes a day.  I usually use the dog's dinner 
or breakfast for cookies.  When that is gone, I stop for the day.  This 
keeps the session short and the dog wanting more.  Don't drill and don't 
train too long.  The next day when I come out, the dog usually has forgotten 
what we were doing the day before and it takes a few iterations before he 
remembers and does the performance as well as the previous day.  This is 
normal so be patient.  Just take a minute or two to let him remember.

After a few days, the dog should be pretty good about putting his feet in 
the proper position.  If not, then you might be rewarding inconsistantly.  
Make sure that you are not giving out cookies for 'close enough' performances.  
If the criteria changes with each performance, you can not expect the dog 
to learn exactly what it is that you want.  Be patient and wait for and 
reward the good performances.  Resist the temptation to physically place 
your dog where you want him to be.  Allow him to experiment with foot 
placement to see if it will earn him a cookie.  If you allow your dog to 
guess and earn rewards for a consistant performance, he will learn quickly 
and give you a reliable performance.

When you get a good performance with the dog in front of you, start 
standing to one side.  Slowly change your body position and location 
relative to the end of the board while still requiring the same performance.  
You want the dog to learn that it is the end of the board that earns him the 
cookie, not being in front of you.  If your dog is having trouble with this 
concept, it is probably because you have moved too far too soon.  Move just 
a few inches at a time so it is not as noticeable to the dog.

When the dog can do the two on two off with you standing beside the dog 
walk, you can now teach him to go up and over it.  If possible, start with 
a dog walk that is low to the ground.  Or lift the dog to the top of the 
down ramp and let him run down to the contact.  He should assume his two on 
two off position.  If he does, give him a cookie.  If he doesn't, don't 
make a big deal out of it.  Put him back up on the board and try again.  
Eventually, keep back chaining until the dog is running from the beggining 
of the board to the end and stopping in the contact zone and waiting for 
your release.  Sometimes release him and let him come to you for a cookie 
and sometimes make him stay there and walk up and give him a cookie.  Keep him 
guessing.  Don't be predictable.  

I have found that dogs learn best when there are no wrong answers.  There are 
only answers that are better than others.  The best answers are ones that 
earn a cookie.  If the dog is afraid to guess because a wrong answer gives 
him an unpleasant result, he will be reluctant to try things for you.  Aussies 
by nature are eager to please and eager to work with you.  If trained this 
way, they find it a game and like guessing what it is you want.  If a dog 
makes a mistake or does not do the performance when you are pretty sure he 
knows it, say "Oh well", don't give a cookie and try again.  The dog might 
be testing you to see if you feel like giving him a cookie for a substandard 
performance.  He's a smart dog for trying but don't fall for it!

Contacts must always be practiced.  Once you have a nice contact performance, 
don't assume that you are done training.  At least once a week, I spend a 
few minutes reminding my dogs that two on and two off is really good and 
they get cookies for it.  I don't let them forget.  If you don't reinforce 
a behavior it will go away.  If you don't pay an employee, he isn't going 
to show up for work.  So pay your dog each week for doing nice contacts.

[Note:  When I refer to the term 'cookie', I mean anything that motivates 
your dog.  It could be his favorite treat or it could be a toy.  If your 
dog likes toys better than food, whenever you reward him, give him the toy 
and play tug or throw it for him.  For a training session, it's best not to 
throw it too far since you don't want him really tired before you are done 
training.  If you have a dog that will take the toy and run off, tie a rope 
or string to it and play tug.  This way, the dog has to interact with you.  
At the end of the training session, or when the dog does a spectacular 
performance, throw the toy and let the dog celebrate by running around 
with the toy.  Let your dog feel like he has won the game.  It will build 
his confidence and make him eager to get out there tomorrow to try the game 
again.]