No Reward Markers

[corvus]

NRMs are, I think, a comparatively safe topic when it comes to discussing training methods. Do you use them? Why or why not?

I started out using one because I thought it would help my dog understand what I wanted from him and decrease frustration. I decided after a while that this was not the case with my particular dog at least. I stopped using it because it did not seem to make any difference to his level of frustration. I also trained an interruptor that meant "stop what you're doing" and used that for a while in training specifically when he was about to break early or something like that.

I have since given up on the former and am weaning myself off the latter. There are three main reasons why. The first one is because I talk too much when I train. I originally felt this was just my style and my animals needed to get used to it. Then I got over myself somewhat and realised that my hare at least actually kind of disliked my constant verbalising. I got out of the habit with him and it carried over to my dog training. With less talking I paid more attention to body language and that seemed to do nothing but help us all.

The second one is that I have come to think that NRMs are a bit of a waste of time. If my reward rate is high enough, I shouldn't really be wasting time telling my dog he needs to try again. He knows he needs to try again because he hasn't been marked and rewarded. I have so far found this to be true in practice, even when my reward rate wasn't really high enough.

The third is kind of two-fold. On one hand, a NRM can become punishing and the last thing I want for my soft, low drive dog is for him to feel like he doesn't want to get a NRM. He would lose interest in training with me. On the other hand, a NRM can be a bit like an interruptor, and if you lean on those things too heavily they lose effectiveness, at least in my experience so far.

In all honesty, I think that dividing myself between "this is what I want" and "that is what I don't want" just makes me fall into the first trap of too much chatter. Concentrating on just what I do want means I'm rewarding more often, which makes training more fun for everyone.

Discuss. ;)

[laffi]

I think NRM is punishing to certain extend... well it's certainly, by definition, is not rewarding!!!

I use it with my highly driven aussie. With my soft golden retriever I try not to use it as she worries about it too much. I do use it more now because she is much more confident, but it took us years (!) to get here.

Anyway I would only use it if the dog was successful at least 80% of the time.

(This reminds me I forgot to answer your PM... better do it now!)

[Cosmolo]

Yes i use them- with my dogs and some clients dogs. Do need to be careful with them though, especially when introducing them. I think they actually help to build confidence when used correctly- even in soft dogs as the dog starts to use a different part of their brain and really has to think for themselves.

[Janba]

I have been using a NRM with Cole in herding training lately and I have found it helps. I am currently training for Advanced A and Inter B so the driving is more complicated than the lower levels.

Herding is a bit different to obedience and agility in that the reward is being allowed to work the sheep and you do have that 3rd factor of the sheep to contend with. You want a dog that obeys you instantly unless the sheep do something unexpected then you want the dog to react instantly. I do use a conditioned reward word when he has done what I want and a NRM when he is wrong.

ETA My dog is usually between 90 and 20m away from me when training so cannot give a physical correction unless I throw my stick at him.

Edited September 20, 2009 by Janba

[The Spotted Devil]

I use it but very sparingly...I find it very useful when Zig is on the verge of really understanding an exercise. It is then he will try and pre-empt the command to gain the reward. A very gentle and calm "ah well" and we try again. At this point I will set him up to succeed if I can and heavily reward the correct response. Too much NRM and Zig will switch off. If you are teaching precise behaviours I think you need something to indicate "try again" that doesn't destroy the dog's confidence. I also used it to help teach the casual recall too. He very quickly learned that there are consequences for not coming when you are called! "Ah well" and put back on lead with no sniffing or peeing or fun!

[corvus]

Cosmolo, could you explain how you would introduce a NRM to increase confidence? I was participating in an interesting discussion recently about the use of pressure in training and I noted that I thought a little pressure served to increase confidence in my hare because if it was low level enough it was a little like flooding and he'd get through it without moving a millimetre away from me and then he'd relax. Release the pressure a little and he'd become a little bolder than he was before. Do you think it's remotely similar?

Also, if a dog is taught to free-shape, is that the same as learning to think for itself as a result of NRMs?

[Cosmolo]

Just wanted to add that with the dogs i use the no reward mark with, i like the frustration and intensity that then comes after it- i don't get anywhere near the same response by simply not rewarding.

[laffi]

I fee shape most of the stuff with my dogs (agility and obedience). With Midge (my aussie) I use NRM markers. With Laffi (my golden) I don't say anything, she would switch off very easily if I used them too much (which is still very little).

To be honest, I consider a NRM a punishment (since it's telling your dog a reward is NOT coming) so I try not to use them too much. I don't consider them kind for the dog.

[Cosmolo]

Corvus, i think i can teach the dog that trying again and really thinking for themself can be very rewarding, rather than them hanging back and waiting to be 'helped' or told in some way if that makes sense. Because they then feel they have more control, their confidence increases as a result.

You need to be careful though as if you don't know when to push and when not to, the soft dog whose confidence you're trying to build will shut down. I worked with a very weak nerved dog who with a NRM to begin with was very very unsure but within one session he was improving and gaining so much confidence in every way- he was more confident with me, less noise sensitive than he was earlier and less worried about quick movements. I believe its because he was using a different part of his brain- which was stressful for him to start because he'd never done it before- but this then created more independence and more confidence as a result.

And yes i think it is very similar to the kind of process the dog goes through when free shaping.

Hope that makes sense!

ETA While technically a NRM is a punisher- i don't think this is true for all dogs once they understand the concept.

Edited September 20, 2009 by Cosmolo

[Rusky]

I use and teach sparingly for different reasons I honestly do waver between 'to teach or not to teach' I find that people who haven't quite grasped what it is and how to use it use a nrm too often without the reward when the behaviour is correct. I find that frustrating so I tend to tell people who I feel need a nrm that it is special, needs to be taught properly, used properly and make sure 100% that they are.

I do find nrm very useful for grab the food type dogs and humpers ;)

[Cosmolo]

Here's another thought to ponder. For the dogs where we say they're too soft for a no reward mark to be used- what about a non verbal NRM- are we using a NRM without acknowledging that we are? eg At times when a reward is removed, is that the same to the dog as a no reward mark? But then the difference becomes IMO- that in those situations the dog is still not learning to use the independent thinking side of their brain as they would with a no reward mark which means to them 'try again' without additional command..

I don't know if that makes any sense.. Its been a long day!!

[Erny]

I think NRM is punishing to certain extend... well it's certainly, by definition, is not rewarding!!!

I use NRM's ..... it is feedback for the dog, just as "good" is my feedback for "you're doing right but exercise not over".

I don't think NRM's are or need to be punishing. I will use a NRM to mark a 'mistake' by my dog, then guide him back to the place he was (eg a broken sit/stay) and recommand. "Good" will follow when he is holding the sit/stay.

My boy responds well to both of these forms of 'feedback'. I think dogs find them helpful if we've been consistent with them and they know what they mean.

I think it would be more stressful on a dog not to receive 'feedback' so he doesn't know WHY he's not receiving an anticipated reward. I also don't think it is necessary to use a NRM and leave the dog to guess at what he's supposed to do, unless you are purposely using the shaping method of training for some specific reason.

Edited September 20, 2009 by Erny

[Cosmolo]

I think that Erny sums up two ways to use a NRM- either when shaping where i then let the dog 'work it out' or as feeback where you then help the dog to do the right thing and earn the reward.

When introducing the NRM i do the latter and then when the dog is ready move to the former. Its the former where i find with some dogs i get ultimate enthusiasm and big improvements.

If giving the NRM, then helping/ showing the dog- whats the difference between the verbal and non verbal NRM? I use NRM's with the aim of then getting the dog to think and try themselves, rather than continuing to help them because then i don't see the difference between giving a verbal NRM and a non verbal witholding of reward. This is assuming of course the dog knows the exercise we're doing.

[huski]

I think NRM is punishing to certain extend... well it's certainly, by definition, is not rewarding!!!

I use NRM's ..... it is feedback for the dog, just as "good" is my feedback for "you're doing right but exercise not over".

I don't think NRM's are or need to be punishing. I will use a NRM to mark a 'mistake' by my dog, then guide him back to the place he was (eg a broken sit/stay) and recommand. "Good" will follow when he is holding the sit/stay.

My boy responds well to both of these forms of 'feedback'. I think dogs find them helpful if we've been consistent with them and they know what they mean.

I think it would be more stressful on a dog not to receive 'feedback' so he doesn't know WHY he's not receiving an anticipated reward. I also don't think it is necessary to use a NRM and leave the dog to guess at what he's supposed to do, unless you are purposely using the shaping method of training for some specific reason.

I agree with the above ;)

I use a NRM and never had any problems with it. I have a confident dog who likes to get feedback.

[Erny]

I think that Erny sums up two ways to use a NRM- either when shaping where i then let the dog 'work it out' or as feeback where you then help the dog to do the right thing and earn the reward.

Either way, it is feedback. The only difference is the training method you use, and that has nothing to do with the NRM having any different meaning, use or bearing.

Edited September 20, 2009 by Erny

[Cosmolo]

You don't think that using it in those two different ways changes the meaning of the NRM at all? ;)

[Erny]

Unless I have mis-read (yes, it's been a long day and I confess to weariness), no I don't. You're still telling the dog "wrong and you don't receive a reward for that", whether you are shaping or using some other training method. It is a marker that signifies there is no reward.

Edited September 20, 2009 by Erny

[Cosmolo]

Hmmm. I agree that the marker says that in both examples, i just think that it also tells the dog to try again themself when they are left to work it out whereas it would not say that if the dog was then helped/ shown what to do.

[Erny]

Hmmm. I agree that the marker says that in both examples, i just think that it also tells the dog to try again themself when they are left to work it out whereas it would not say that if the dog was then helped/ shown what to do.

It is the method that you use which is the difference. Not the NRM.

[Cosmolo]

I realise its the method that creates the differing response- but doesn't that then mean that the NRM means different things to a dog trained with method A than it does to the dog trained with method B? ;)