Dog Forum banner
Status
Not open for further replies.
41 - 49 of 49 Posts

·
Premium Member
Joined
·
11,875 Posts
So don't want to get too OT and also haven't read every word of this post but...

The issues being mentioned with food (ie. dog checking for food) as a reward are actually indicative of training errors on the handlers part, not actually the usage of food. ;)
http://www.dogforum.com/training-behavior-stickies/thoughts-training-food-1219/

The tugs on a lead or leash corrections are absolutely not a positive reinforcement based method. If it actually works you're using positive punishment and negative reinforcement (dog is working to avoid something unpleasant).
http://www.dogforum.com/training-behavior-stickies/4-quadrants-operant-conditioning-23702/
If having to give multiple tugs per session each time out and about, the tugs are not effective. What happens is that you're basically nagging the dog. Might momentarily stop the dog(reinforcing the the handler as it appears to work), but doesn't really overtime. Something else that often happens is that the dog gets used to the corrections so the owner ends up increasing the intensity. The dog get used to that, and the owner increases intensity again. So on and so forth.

We're talking politely walking on lead not really ''heel'' (a focused heel like one sees in various sports) right?
"Heel" is actually very complex and a difficult behavior for dogs and humans. So much is involved... a precise position in relation to the human, head up and eyes on the handler at all times, the dog learning body awareness and how to move it's body to maintain position, speed changes, direction changes, pivots, autosits (and needs to be a tuck sit to maintain position-most people teach a rocking sit), training duration and distractions, etc. And that's just the dog's part! Doesn't even touch on the handler. Heelwork is also a very time consuming and expensive behavior to both train and maintain. Because of the difficulty, a happy, enthusiastic, and precise focused heel takes an incredible amount of higher value rewards. Praise alone isn't going to cut it for the vast majority of dogs.

When it comes to polite walking, I still don't find praise alone super effective for a great many dogs. Food and toy/play ime is very effective and what I use to begin training.
After the concept is loosely understood, I begin switching to access to the environment/distractions (sniffing, exploring, greeting people, etc.) as the reward for a loose leash. Along these lines:
Pulling results in freezing or walking away (yeah, results in tension on lead but no need to actually tug/pop) and a loose lead results in gaining access to what the dog wants.
Still in new and/or very distracting locations I use food and toy rewards for walking short lengths on a loose lead. As we walk along I can increase distance between rewards and by the 3rd time visiting that location, I generally so not need to use much food at all.

Also, food and toys serve as a reward (even for my mature adults) for excellent behavior when the distraction cannot be used as a reward. For example keeping a leash loose when a critter runs away, walking nicely past a barking dog, sticking with me when a biker, jogger, or skateboarder zips past us, etc.)

Last, ime, trainers who are generous with their reinforcers and using high value reinforcement tend to actually have the easier time when training for distractions. Generous trainers get in far more successful repetitions. That equates to quicker leaning not only because of the reinforcement history of the behavior but because the dog is likely to make fewer errors during the initial training process. Also, the dog enjoys being near and interacting/working with the handler. When it comes to dogs belonging to handlers who do not use a decent amount of higher values reinforcement, they are often less engaged in interacting/working with the handler and the environment trumps working with the handler.;)
 

·
Registered
Joined
·
94 Posts
Discussion Starter · #42 ·
I have never found it to be harder honestly, for me at least it seems to be easier. I think it probably depends on whether or not the dog has already been trained using treats. I kind of came upon this training method by accident, I was training a friend's dog to stay and I didn't use treats for this, well amazingly where it had taken probably about a week for the dog to learn sit and such commands when using treats, it actually only took two days for the dog to learn stay, and I was even more amazed when I learned by using the command that the dog obeyed that command without fail whereas with the others the dog would only obey some of the time. Right then and there I stopped using treats and switched to this new method.

If you wish to use my method then I would just start using it for everything now because if the dog gets spoiled on treats it will make it hard to train it the other way.

That is quite possible, I am currently working with a friend's dog on heel and the dog will start to try and go faster when going down hills and up hills. If you are able to, try and increase your pace a bit to help the dog, just make sure he stays at your side.
That book...just glancing at the amazon page, is it tailored for Retrievers or is that just the breed the author used?

I can try to increase my speed. However, he usually increases his. It becomes more a test in arm strength than anything!

So don't want to get too OT and also haven't read every word of this post but...

The issues being mentioned with food (ie. dog checking for food) as a reward are actually indicative of training errors on the handlers part, not actually the usage of food. ;)
http://www.dogforum.com/training-behavior-stickies/thoughts-training-food-1219/

The tugs on a lead or leash corrections are absolutely not a positive reinforcement based method. If it actually works you're using positive punishment and negative reinforcement (dog is working to avoid something unpleasant).
http://www.dogforum.com/training-behavior-stickies/4-quadrants-operant-conditioning-23702/
If having to give multiple tugs per session each time out and about, the tugs are not effective. What happens is that you're basically nagging the dog. Might momentarily stop the dog(reinforcing the the handler as it appears to work), but doesn't really overtime. Something else that often happens is that the dog gets used to the corrections so the owner ends up increasing the intensity. The dog get used to that, and the owner increases intensity again. So on and so forth.

We're talking politely walking on lead not really ''heel'' (a focused heel like one sees in various sports) right?
"Heel" is actually very complex and a difficult behavior for dogs and humans. So much is involved... a precise position in relation to the human, head up and eyes on the handler at all times, the dog learning body awareness and how to move it's body to maintain position, speed changes, direction changes, pivots, autosits (and needs to be a tuck sit to maintain position-most people teach a rocking sit), training duration and distractions, etc. And that's just the dog's part! Doesn't even touch on the handler. Heelwork is also a very time consuming and expensive behavior to both train and maintain. Because of the difficulty, a happy, enthusiastic, and precise focused heel takes an incredible amount of higher value rewards. Praise alone isn't going to cut it for the vast majority of dogs.

When it comes to polite walking, I still don't find praise alone super effective for a great many dogs. Food and toy/play ime is very effective and what I use to begin training.
http://www.youtube.com/watch?v=0mzSgtAuHyU
After the concept is loosely understood, I begin switching to access to the environment/distractions (sniffing, exploring, greeting people, etc.) as the reward for a loose leash. Along these lines:
http://www.youtube.com/watch?v=QawGibMta54&index=1&list=WL
Pulling results in freezing or walking away (yeah, results in tension on lead but no need to actually tug/pop) and a loose lead results in gaining access to what the dog wants.
Still in new and/or very distracting locations I use food and toy rewards for walking short lengths on a loose lead. As we walk along I can increase distance between rewards and by the 3rd time visiting that location, I generally so not need to use much food at all.

Also, food and toys serve as a reward (even for my mature adults) for excellent behavior when the distraction cannot be used as a reward. For example keeping a leash loose when a critter runs away, walking nicely past a barking dog, sticking with me when a biker, jogger, or skateboarder zips past us, etc.)

Last, ime, trainers who are generous with their reinforcers and using high value reinforcement tend to actually have the easier time when training for distractions. Generous trainers get in far more successful repetitions. That equates to quicker leaning not only because of the reinforcement history of the behavior but because the dog is likely to make fewer errors during the initial training process. Also, the dog enjoys being near and interacting/working with the handler. When it comes to dogs belonging to handlers who do not use a decent amount of higher values reinforcement, they are often less engaged in interacting/working with the handler and the environment trumps working with the handler.;)
I'll have to peek at these videos later on! I'm curious to see what others in this thread think about this as well (should they opt to opine). Didn't think I'd have two schools of thought, much less three!
 

·
Registered
Joined
·
94 Posts
Discussion Starter · #43 ·
So don't want to get too OT and also haven't read every word of this post but...

The issues being mentioned with food (ie. dog checking for food) as a reward are actually indicative of training errors on the handlers part, not actually the usage of food. ;)
http://www.dogforum.com/training-behavior-stickies/thoughts-training-food-1219/
You know, I see video after video of dogs paying more attention to the trainer as opposed to the food. Maybe I just need to tire him out before the food starts. I have managed to hide treats before, but I have to really hide my hand digging into my pocket from him.
 

·
Premium Member
Joined
·
11,875 Posts
Are you familiar with using a clicker or a marker in training?
Done right, Ime that's the trick!

The click or marker (often a tongue click or short word like "yep!") happens when the dog does the right thing (''marks'' or indicates the correct choice and tells the dog a reward is on it's way).

To use a clicker or marker correctly you want to train yourself to have a sort of 4 beat pattern. Mark, pause, reach for treat, feed.

With capturing, targeting, and shaping generally you'll be keeping food and toys out of the dog's sight (luring requires food in hand at first but done correctly you'll be removing food in hand early on). The dogs catch glimpses of the toy and can of course smell the food. They know you have goodies. BUT that four beat pattern ensures that its your marker/click that is important and meaningful. If done incorrectly or training without a marker, the dog is looking for the reach or hand movement as that's what is indicating the correct choice and delivery of a reward vs. watching the handler and listening for the marker.

Does that make any sense?

This is a nice vid for people new to clicker training. I don't do everything exactly the same (I do not load the clicker, I just get to work with a simple exercise) but def good info in there!
 

·
Premium Member
Joined
·
11,875 Posts
Oh! A couple other things that really help in the long run....

Once the dong understand and is reliable about doing the behavior when asked, begin to vary the reward.

Set attention/eye contact as a criteria for the behaviors you train. So for example, for my dogs ''sit'' doesn't just mean butt on the floor. It means ''put butt on floor and keep eyes on me''.

And just a trick to really jumpstart focus on your face. Spit food rewards. Or mark and then take a treat out of your mouth feeding from hand. Though you generally only want to do this with stuff that requires the dog to face you. Maybe heel position if your dog is large enough to make eye contact while next to you and in that case you would make and feed from left hand.
 

·
Registered
Joined
·
94 Posts
Discussion Starter · #46 ·
Are you familiar with using a clicker or a marker in training?
Done right, Ime that's the trick!

The click or marker (often a tongue click or short word like "yep!") happens when the dog does the right thing (''marks'' or indicates the correct choice and tells the dog a reward is on it's way).

To use a clicker or marker correctly you want to train yourself to have a sort of 4 beat pattern. Mark, pause, reach for treat, feed.

With capturing, targeting, and shaping generally you'll be keeping food and toys out of the dog's sight (luring requires food in hand at first but done correctly you'll be removing food in hand early on). The dogs catch glimpses of the toy and can of course smell the food. They know you have goodies. BUT that four beat pattern ensures that its your marker/click that is important and meaningful. If done incorrectly or training without a marker, the dog is looking for the reach or hand movement as that's what is indicating the correct choice and delivery of a reward vs. watching the handler and listening for the marker.

Does that make any sense?

This is a nice vid for people new to clicker training. I don't do everything exactly the same (I do not load the clicker, I just get to work with a simple exercise) but def good info in there!
http://www.youtube.com/watch?v=_wv1uvvqaSw
I had started clicking with my tongue at some point in training. It was remarkable how receptive to it he was. I think I abused it a little and he no longer comes bounding over from across the yard. However, that 4 beat thing makes a lot of sense. I mean, it's obvious-you make the click, you give them a treat-I just never put it together to make the clicking the marker for whatever reason. I think I always tried to rush things so I fed him as I clicked. Going at that 4 beats might make things easier.
 

·
Premium Member
Joined
·
11,875 Posts
Oh absolutely!
If the click doesn't happen first it has no meaning! (or at least not the meaning needed for clicker training)
And it's one of the more common mistakes made by people trying out clicker training.

A second common mistake is that people tend to use the click as an attention cue, and that's not what it should be. The click should not be used to tell the dog what to do (that's where your hand and verbal cues will come into play). It's intended to tell the dog it responded correctly and that he earned a reward. You want to click the action you wish to reward. So for recalls, you'll call your dog, click as your dog is on his way to you, and then reward once he's with you. You won't click to try and get him to come to you. :)
 

·
Registered
Joined
·
94 Posts
Discussion Starter · #48 ·
Oh absolutely!
If the click doesn't happen first it has no meaning! (or at least not the meaning needed for clicker training)
And it's one of the more common mistakes made by people trying out clicker training.

A second common mistake is that people tend to use the click as an attention cue, and that's not what it should be. The click should not be used to tell the dog what to do (that's where your hand and verbal cues will come into play). It's intended to tell the dog it responded correctly and that he earned a reward. You want to click the action you wish to reward. So for recalls, you'll call your dog, click as your dog is on his way to you, and then reward once he's with you. You won't click to try and get him to come to you. :)
I totally began to use it as an attention cue and after a day or two it lost it's appeal for him.

What I envision is calling him from across the yard. He looks over or starts walking. Then I click and he runs. I should probably have a treat for him?

Last night I brought him into the bathroom for some silk training with very little outdoor exercise. I actually started the 4 beat thing and as soon as I clicked he started looking for a treat. Hopefully tonight I can get him in a bigger room.

What hasn't helped recently is that my wife has started to join us with our newly acquired 8 month old chuhuahua. So in addition to being particularly excited about going on the walk, he is also constantly looking back to ensure his new activity buddy is right there.

I have another thread here where I'm considering a front clip harness or gentle lead. This is where I try to merge the posts a little and ask if either would help him keep at heel and get trained to heel. Sometimes command and reward just gets tedious...
 

·
Registered
Joined
·
94 Posts
Discussion Starter · #49 ·
So he is doing better on the leash and staying with me. I haven't done much silk leash training (may get back to it tonight). However, we recently adopted an 8 month old Chuhuahua. Between training days for him, he gets to come for a walk as well (instead of running around the house or yard). Here's the thing...Alfred-the one I've been working on-regresses behavior wise when joined on the walked by our newest addition. Will he just straighten out with enough walks with both of them? He pulls for the majority of the walk if we have them both. Will I just have to walk both of them enough times to reduce the excitement or is that another way? Would a gentle lead help drive that behavior faster?
 
41 - 49 of 49 Posts
Status
Not open for further replies.
Top