Quote: michael_wise

Could this be remedied with further training and better timing with rewards?

I can only wish that my training problems were my dog offering too many behaviors.

Sure, any behavior can be remedied by better training. But it makes sense to have a marker for the WRONG thing as well. When you mark the wrong behavior and withhold the reward, it makes the pup work harder for it because it builds frustration. Frustration makes for better and quicker responses when done and used correctly in training.
As for better timing - well, when you mark/treat, the two have to be separate. You first mark (yes or whatever) then you treat. This makes it easy to identify. Dogs follow physical movement quicker than voice, so when you mark and treat at the same time or only treat, the physical act of digging the treat out of a pocket or bag, or just holding the treat in hand makes the dog watch THAT (bag, hand etc) because you are actually marking the behavior with the movement of the hand. Also, by keeping them separate, the dog knows that the payment is coming. So you don't have to have it in hand, or have the dog next to you. The marker (click, yes, etc) is basically like a release, and the dog can then expect a reward. So on a down stay from a distance, saying "yes" will release the dog, and it will come running to you for the reward it knows to expect.
Make sense?

Yes. A properly trained dog using this method doesn't need a treat for every behavior once it's solidified and been given a name, in fact commands are responded to with much more enthusiasm after this point if the rewards are infrequent. Generally a dog will offer new behaviors only when the trainer is actually paying attention, otherwise what's there to reinforce it? I'm not sure what was going on with the dogs at the seminar; sounds like there was either a 'piece' missing or they just weren't finished yet, but even that doesn't sound right. Even an unfinished dog shouldn't be 'offering behaviors' as an attention seeking kinda thing, unless it's been reinforced by someone who didn't know that they shouldn't, and a dog trained to sit or down should remain that way until told otherwise. It sounds like the method may have been misapplied?

Reward doesn't have to be food every time, and you don't have to mark every time with an older dog.
YOu can do 4or 5 commands in a row, then end the exercise with a marker and reward (which can also be playtime with a toy)
The point is, when you say the marker word, or a click, you should ALWAYS pay the dog.
No this isn't used out of context, and if you ever have a chance to watch one of these seminars, you will really be impressed with the dogs trained in this method. It works very quickly, thoroughly, and reliably.
When you string commands together, you build up the frustration for the reward, making it more desireable.
The dog knows that you will pay it, it just doesn't know when.

Offering behaviors happens in a dog that has not been corrected for offering behaviors in the past. It's when the dog is punished for trying to figure out the problem that they stop offering behaviors and wait for you to show them what you DO want so they won't get corrected again. It's a great way to reduce drive.

 Quote: Cameron Feathers

Make sense?

Yep!

 Quote: Cameron Feathers

Offering behaviors happens in a dog that has not been corrected for offering behaviors in the past. It's when the dog is punished for trying to figure out the problem that they stop offering behaviors and wait for you to show them what you DO want so they won't get corrected again. It's a great way to reduce drive.

This is what I did to my dog pre-Leerburg. End product was a dog that took no initiative during training.

Over coming this has been difficult for my limited experience, but we are doing better.

Thought a little more about this:
Basically, this was illustrating the need to have a positive marker, and a marker for when the dog offers the wrong behavior. It is not a correction, and eliminates the need to correct the dog when training unless it's for a rank or aggression issue.
By training this way, the dog stays engaged, and any time it makes a mistake, or misunderstands you, it doesn't get corrected for doing it wrong. A finished dog/older dog SHOULD be offering behaviors when you don't respond, or if you mark the behavior with an upbeat "nope" or something similar. It keeps the dog interested, and makes it try to figure out what you want. The dog I was talking about was finishing a recall, and because the incorrect response was not marked in any way, but the reward was withheld, the dog kept trying other things to get the reward. By marking the incorrect behavior, it makes it easier for the dog to understand that you want the dog to keep trying, and it stays in drive. Then, when the dog DOES do it right (even with an older or "finished" dog, the behavior should be marked and rewarded.
Even older dogs need continued positive reinforcement. YOu don't need to reward EVERY behavior, but you should still be rewarding frequently.

I re-read too and realized what you meant.

Instead of having a "negative marker", for the lack of a better term, I was correcting my dog. This took enthusiasm out of his training that we are finally getting back after forever.

He wouldn't shut down or really stress out, but only tried what he knew worked in the past and NEVER, I mean NEVER offered anything new. Tough to train that way.

We're on the same page now. More often than not, I think people are talking about the exact same thing, but word stuff so differently they think the other person is off a little.

I want to make sure I understand what you're saying, Cameron, and that I'm not doing the wrong thing.

For example, when I ask for a finish True will sometimes get a little sloppy with his position, usually when he's all fired up. I do a sort of throat-clearing "AHHH" sound, which he seems to understand is "Nope, that's not what I want."

He'll then try it again...and sometimes he needs a second "AHHH" before he nails the correct position and gets marked and rewarded for it. He sometimes whines in a frustrated/excited way when he's asked to do it over.

Am I using a negative marker in the right way?

With commands he knows 100% he'll keep doing that one command until he gets it right. With commands he knows less well he'll start offering up other things...which gives me a clue that he hasn't fully learned the command I'm asking for.

Does this make sense and am I reading this correctly?

yes, Sarah. The dog is still actively engaging and offering up behaviors when you withhold the reward.
The word you use doesn't matter, as long as it is not something that you use when, say, you catch the dog in the trash, or use a verbal correction. The marker that you use has to be only reserved for marking behavior that you don't want. It has to be upbeat (which it sounds like you are)basically, you are saying, nope, try again! to the dog. Then as soon as the dog offers the RIGHT behavior you mark it and treat it.
The biggest thing with markers is to make sure that the reward FOLLOWS the mark, not at the same time.
The concept goes back to Pavlov's conditioning - when they rang the bell, the dogs started salivating in anticipation of the food that was coming. It became an autonomic response. This is "classic conditioning" The timing of the reward is one of the hardest things to get. It almost seems automatic to reward at the same time.
When you are trying to use this method, well, it's a lot easier to lure as well as mark/treat. Don't use the command with the lure until you KNOW the dog will be lured into exactly the right position. Then you put the command first, lure, mark, reward. Eventually the dog will start offering the behavior after the command and before the lure, to get the reward faster.
Once that happens, you are ready to start using the "nope" or whatever you use. Even once the dog is finished, any time they are actively engaging, but say, at the heel, the dog forgets the automatic sit. Rather than a correction, you just say "nope" then when the dog offers the behavior, you mark/reward. Basically you can mostly eliminate the correction in OB work altogether. At least in the traditional sense.
Hope that helps. Sounds like you have a really good understanding of it already.

Yes, that does help...Thank you, Cameron!