Quote: david c.frost

The reward at source (ball, tug, etc) in my opinion, is in fact a bridge. Rewarding at source, initially, teaches the dog to go to source because that is where the reward comes from. As you pointed out, with many dogs, it's the play aferward.

DFrost

Much of what is "rewarding" is based on the selection testing and early training.
I am a fan of a system that instead of worrying about rewards (a way overused term in dog training) we try to capitalize on the attributes of the type of dog selected.
Tug-O-war is for some dogs great, it is the physical play and interaction that they need. I prefer a dog that in its mind is flushing its prey. It is hunting and flushing its prey. It is not so dependent on a "reward". Here is the example I use: many many dogs hunt and chase rabbits. In my experience they catch few if any. This is no reward system and yet it is rewarding, the dog just feels good doing it. This approach diminishes the handler involvement in the find and indication to a minimum. Here a tug-O-war game may not be important at all. My previous dog I did not play tug with at all. He was hunting and trying to flush what he viewed as a prey item. My job as handler was two fold. Give him areas to hunt in and keep him from hunting anything other than what was assigned. Pretty simple system tha is quite robust in nature. It goes back to the idea that the more natural a behavior is the more reliable it will be.
The down side is that many dogs cannot work solely this way. Particularly retrievers who work more out of hunt/play than hunt and prey. This is not a hard and fast rule as all breeds overlap considerably in their behaviors. The onset of this system is also tedious in setting up aides.
So, back to tug-O-war. It may not be necessary for the dog at hand and for others imperative. It depends on training and selection testing methodology.

So how would you make that determination?

The dog works if you just throw a ball at the source; I know some retrievers will solicit a game of fetch by throwing the ball at you. He does not, but if you throw it he will chase it. In my human mind, that is not "much of a reward" but maybe it really is enough. I gather you may not even throw the ball?

The tug pulls him back to me and I can get the toys and it "seems" to jazz him up. I take the toys away and he starts searching again.

I have heard that dogs that have run deer a few times quit because they never catch it and that a behavior not rewarded extinguishes...........but then again..........isn't the reward the completion of the hunting sequence? Most of the time the wolf does not catch its prey but it has to be enough of the time for the wolf to keep doing it (and not starve).

You are messing with my paradigms, I like this kind of discussion.

Like I said in my first post. With some dogs, tug-of-war is the reward. With others it may be the simple chase and retrieve. Labs we purchase from serious duck dog breeders will rarely play tug. With them it's the retrieve. To the trainer, it's important to identify which will be powerful enough to drive the behavior you are soliciting.
As for your wolf analogy, that really doesn't compare. There is a significant difference between a dog that will play tug of war, hunt and retrieve all day and survival.

 Quote: kevin sheldahl

 Quote: david c.frost

The reward at source (ball, tug, etc) in my opinion, is in fact a bridge. Rewarding at source, initially, teaches the dog to go to source because that is where the reward comes from. As you pointed out, with many dogs, it's the play aferward.

DFrost

Much of what is "rewarding" is based on the selection testing and early training.

Of course it's based on the selection testing. I wouldn't select a dog for training that didn't have any prey drive or desire to hunt. While I don't disagree that one should "capitalize on the attribues of the type of dog selected" I do have the luxury of making the selection. To that end, I select the dog that has the attributes I'm looking for.

DFrost

 Quote:

It goes back to the idea that the more natural a behavior is the more reliable it will be.

 Quote:

Particularly retrievers who work more out of hunt/play than hunt and prey.

I agree with this totally, (except with the retreiver part )


A dog that is searching not because it is fun, but because he is HUNTING from that basic desire is preferable. That said I would not go without rewards for what I do. Hunting must be rewarding, but getting the "kill" (at least sometimes) is important too IMHO.

 Quote: nancy jocoy

But if it is the brief ensuing game then to me swapping the ball for a tug or playing two ball is the same thing; the reward is not happening right at the source but around it and not immediately at the time of the find.

So is the ball "popping out" of source the actual reward or a bridge like a clicker and the reward is the play?

The only thing I would worry about with the two toy system is that eventually the dog anticipates that a more fun thing follows the first thing (ball first, not as rewarding as tug that dog likes better and can play tugs better with).

Could it happen that the dog would eventually just start looking at you instead of source, anticipating the better toy and tug session, wanting to skip the ball on the source part?

A lot to think about and a lot to watch for............maybe I am overthinking this as I tend to do.

Interesting your mention that stare at source thing. When I got him he had a very nice sit and stare at source alert (was trained as a demo dog using a ball-ejector system) - so solid he would stare at source even as plastic bottles were tossed onto his head.

Now I have no ball ejector, and my aim is not always perfect, but he does sit, momentarily look at source, then looks at me, but stays put. The person I was working with did not have a problem with this; you can tell pretty much where source is by reading the body language going into it.

We have, with a different dog that left source anticipating the reward from its handler and leaving source, had a 2nd person throw the ball at source so the dog did not know where it was coming from and would not cue into the handler. We mix it up and use this with the other dogs too.