Lagging In Heel Work

[Tali]

Hi all

Can anyone suggest ideas to try with a dog that walks beautifully at heel on lead (never a pull, right where I want her) and then lags when she is off lead (but about 1/2 to 1 metre... sometimes gets distracted and I lose her in the about turns).

I've tried food, but when the food is not in my hand, she lags again (not great for trialling!)

I've tried running faster to pick up her pack instinct, but she is happy about 2 metres behind me and doesn't seem to need to catch up.

Put her back on lead to use the check to correct the bad behaviour and she doesn't do it... argh! I talk and call her up but again she is happy back there behind me...

Sorry -just feeling frustrated and running out of ideas...

[Nekhbet]

keep the lead on for all training sessions and dont let her lag at all. Instead of a correction just give her little tugs to encourage, not punish.

do it in short sessions, make it interesting for her as well dont stick to a consistant walking pattern so she becomes bored. Make it fun and exciting, heaps of verbal praise when she does it right not just food, let food be the jackpot for performing.

have you tried using a clicker to mark her behavior?

[Staranais]

I'd probably start again, teaching heel offleash as if it was a brand new exercise. I'd want to keep her excitement and motivation level high with a reward she really likes, special food or a toy, but make her responsible for finding and keeping the heel position to earn the reward. So I'd probably teach her to play the "find heel" game, or practice just a few steps at a time, then mark the behaviour and release her to the reward.

Like Nekhbet says I'd do lots of short sessions to keep her excitement high, and only gradually build up how long you expect her to heel for, only raising the criteria when she's both keen and accurate in her heeling.

I taught competition heeling as a completely offleash exercise from the start with my last dog. I figured if he wouldn't stay with me offleash, then there were problems (with either his motivation or his understanding of the exercise) that sticking a leash on wouldn't really fix (it might cover the problems up like a band aid for a bit, but that's all!) I think I'll do it that way with my next dog, since it worked really well for us.

[laffi]

I would get rid of the leash... work on stationary heel position (rewarding heaps and heaps) make one step reward the dog for staying in position (moving with you).

Don't be scared to use the treats but don't lure her: keep the treats in your right hand, reward in position with your left hand.

Edited June 16, 2009 by laffi

[Kavik]

I agree with Staranais I have been doing the find heel game with a clicker and mark when the dog is in heel position and throw the food. Diesel gets as enthusiastic with this as he gets I start off moving (tried stationary first but he got bored and started sniffing the ground and wandering, if I am moving he is more interested) and then stationary so he had to scoot his behind around to get into heel position. I may teach it this way from the start next time

[Tali]

Thanks for replies so far - I do have a clicker but haven't really used it as yet. Am I right in understanding that I 'load' the clicker with rewards (eg. food). Then when she IS in position, I click (but don't reward EVERY time)?

I have been having trouble with timing the reward as when she does get into correct position (through luring) i then tend to stop and reward her (so she is probably not associating the reward with the behaviour I want...). I guess the clicker would help with this?

I should add that I don't really "punish" with the check chain correction - she is a very submissive dog and I only need a light hand to correct her... I sometimes wonder if she is running behind out of submission?

[Nekhbet]

start her in the heel position and as laffi said take a step foreward, lure but pull your hand slightly further foreward then your knee to make sure she stays in position YES , SIT treat

link it as an exercise and do it step by step she gets the reward after the fact

[Staranais]

I have been having trouble with timing the reward as when she does get into correct position (through luring) i then tend to stop and reward her (so she is probably not associating the reward with the behaviour I want...). I guess the clicker would help with this?

Yes, that's right. It will help her understand what you want, since you can mark the exact moment she does the good thing, and then give her the big reward.

If you just use treats without marking the behaviour first, she'll be confused exactly what earned the reward. You can also get a dog that "sign tracks" (spends more time watching for the reward instead of doing what she's supposed to do, since she's never sure when the reward will appear). Using a marker like a clicker helps to correct that problem as well.

I should add that I don't really "punish" with the check chain correction - she is a very submissive dog and I only need a light hand to correct her... I sometimes wonder if she is running behind out of submission?

I'd guess that she doesn't really understand what "heel" means. If you're across the other side of the room and tell her to "heel", could she find heel?

I'm also thinking that she's probably unenthusiastic about the reward you're offering. Make it something really special, reserve it just for heeling if you have to. Your dog should be excited when you tell her to heel, since she should know she's got a chance to get a really super reward.

I'm A-OK with using a correction collar to teach loose leash walking, but don't think it has a place to teach competition heeling, at least not for the vast majority of dogs.

JMO.

Edited June 16, 2009 by Staranais

[dogdude]

I

taught competition heeling as a completely offleash exercise from the start with my last dog. I figured if he wouldn't stay with me offleash, then there were problems (with either his motivation or his understanding of the exercise) that sticking a leash on wouldn't really fix (it might cover the problems up like a band aid for a bit, but that's all!) I think I'll do it that way with my next dog, since it worked really well for us.

Yes, agree with that.

If your dogs motivation wains with a lack of the sense of food on you, ..simply,.....get it off you, and send it to a food target at various distances. Do not do block heeling with turns, only very large circles both directions.

When dog is fantastic at that, train turns using very little heelwork, then marry it all together later.

Trial training is all about breaking the training down into minute part exercises, and only marrying them together occasionally. Forget the rubbish / routines that go on of a Sunday Obedience class.

Edited June 16, 2009 by dogdude

[Nekhbet]

it depends how much you rely on the leash

schutzhund trainers love the leash as it gives you greater control and you have the power to limit or even prevent mistakes. It doesnt have to be heavy just a fine one on a flat collar to use as guiding tools if you have to.

[Staranais]

Why would I want to limit or prevent mistakes when training competition obedience? I'd rather let the dog make the mistake and mark it as a mistake (immediate NRM and start over with no reward). That way the dog can learn that the mistake is unrewarding, and he then knows that it's not worth repeating.

Preventing the dog from making mistakes might mean the exercise is learned more quickly, but letting the dog experiment a little to find out which heeling variations get rewarded and which don't leads to the dog ultimately having a truer understanding of the exercise, IMO.

Lots of ways to skin a cat and all that, but I just don't see the advantage of using a check chain when teaching a competition heel to most dogs, over teaching it offleash?

[Nekhbet]

i didnt mean to use a check chain with the leash

as you can see from above post off leash and not working can build frustration in the handler, you can easily still mark the dog trying to do something wrong you just dont give it an opportunity to make a big mistake if you get my drift.

why make things harder for yourself? If you have the chance to prevent major mistakes and also have a tool to help guide the dog into position or use it as an extension of your own arm why not?

we saw it at the Uta Bindels seminar, she was mortified people went straight to no leash and then had a list of problems or 'needs correction' a mile long.

experiment a little to find out which heeling variations get rewarded and which don't leads to the dog ultimately having a truer understanding of the exercise

I see your point but it can be frustrating as well to a dog that keeps getting NRMs for wrong behavior.

Edited June 16, 2009 by Nekhbet

[Staranais]

Ah I see now.

I've dabbled in both ways, and I did get far better results teaching heel offleash - but then again I'm pretty hard to frustrate! Doing it that way may not be everyone's cup of tea I suppose.

Of course, if a person has to use a leash merely to stop the dog from running off then I really do think that they need to work on motivating the dog better rather than tying him up to stop him leaving.

Sounds like a good seminar though, I would have liked to have been there to see Uta's method first hand.

Oh just saw your edit.

Yup, it can be frustrating for the dog to get too many NRM, although I have also found that some frustration also builds drive for the reward, so I don't see anything wrong with allowing the dog to fail sometimes. But if you're giving too many NRM and the dog's getting overly frustrated, then I'd simply say simplify the exercise. Start with a smaller "step", one that the dog can be successful at, and only move on when the dog's got the hang of that one. Assumed the OP knew that but perhaps I was wrong - sorry OP if you were confused!

Edited June 16, 2009 by Staranais

[trifecta]

Sorry to hijack your thread Tali! I am actually having the same problem with my Groenendael. She has come through her classes with A grade performances, but now we are up to Novice competition, she is seriously sick of "drill" heel work & doesn't want to work at all unless she knows I have food in my hand. Sometimes she looks like she is just shutting down & it is frustrating because I know how good her heel work can be! I have always been careful not to overtrain her & never ask for more than 3 minutes heel work at a time either! I am a fool really as she has always been ahead of her class & I have stuffed around & taken far too long to get her up to competition level.

I have got hold of a copy of Dawn Jecs "Choose to Heel", have only briefly dipped into it as of yet, but she too uses the reverse theory with the dog off leash or on a long line as the foundation of precise & enthusiastic heel work, & starts with walking anti-clockwise circles & treating the dog when it chooses to assume the heel position by targeting the handler's hand. She builds up to extended rhythm patterns of circle, oval, figure eight & spiral rather than drill heeling, much more appealing to dogs who do not like repetition.

I will definitely be training like this with my next dog, but that is years away! Meanwhile going back to the beginning is a bit daunting, but looks like I don't have much choice

[dogdude]

why make things harder for yourself? If you have the chance to prevent major mistakes and also have a tool to help guide the dog into position or use it as an extension of your own arm why not?

we saw it at the Uta Bindels seminar, she was mortified people went straight to no leash and then had a list of problems or 'needs correction' a mile long.

I guess if you want a robot that cant think for itself you would train with a view of not making mistakes. Pattern training can cause just as many problems as it tries to avoid IMO.

When the pattern changes unexpectedly, the dog gets lost.

A lead is fine, but drive is the upmost important trait that you want in trialling, and any hint of aversive methods will affect drive in negative way to an extent.

NRMs are fine too, as long as you keep them as a feedback tool and not a correction, in a balanced way.

Edited June 16, 2009 by dogdude

[Tali]

Ahhh - I like the ideas of the circles, ovals etc - will try to combine that with clicker and just focus on getting her in the right place before re-introducing the obedience turns etc..

I am aware of the need to use small steps and go backwards when need be, but just couldn't work out what 'backwards step' came between lead on and lead off. The suggestion that I think of off lead heeling as an entirely different skill to on lead heeling is quite a new and useful one - at least gives me something else to try so thank you all!

I think the problem may stem from me not clearly marking the EXACT behaviour I am after. We have got through CCD (scores of 98, 99 and 85 from 100), and I'm about to enter a novice sweepstakes here in QLD, but my girl is not heeling as well as Id like and I want to sort it now rather than have the problem persist into higher levels. I'm not after a robot - just want to work out how to keep the fun in it for Jess as well as me.

I trained my old Kees 10 years ago and he heeled beautifully off lead (never really trialled him - just did it for fun). I was lucky last time though, as my old dog hung at my side like a shadow cos he was VERY food motivated and also scared to be far from me. When he went from on lead to off lead I just took the lead off and away we went... so I'm finding now I don't really know what to do when that doesn't happen! I feel I haven't got my girl's motivation sorted... but she is very smart so it may be that I'm not accurately letting her know what I want,

If I call her to heel across the room, she will trot over and fall in with me, but in her 'lag' position.... she knows what to do, but that suggests to me that she understands 'heel' to be 'walk about 1/2 metre behind me'. sorry if I'm sounding a bit muddled; I'm a) thinking out loud here in the presence of people who have more experience and knowledge than me and b) at work (shhh - dont' tell!)

Dummies question here - what is NRM???

[Staranais]

NRM = no reward marker. Sorry for the jargon!

A NRM is a word that signals to the dog that he hasn't earned a reward this time. The advantage of using a NRM (instead of just not giving the dog the reward) is that it pinpoints the moment the dog did the "wrong" thing and lost the chance to earn the reward. Makes it much easier for the dog to work out what's going on. So it's sort of like a clicker, but in reverse.

It's not supposed to be an inherently aversive sound - it's not a growl, or telling the dog off. It's just supposed to be information, given in a neutral, calm tone.

Some people don't like using a NRM, but I find it really enhances communication with the dog & wouldn't be without one.

[Cosmolo]

I like a NRM too- find it builds drive in my lazy dog who has come along in literal leaps and bounds since i started playing find heel position games off lead with high value rewards, not lures. Trifecta- you'd be surprised at how quickly 'going back to the start' can change the dogs attitude, i didn't find it to be a long process and i had a dog who almost hated heeling at that point, despite the rewards i was using.

[Tali]

Nekhbet - not ignoring your advice re lead on again, but can't seem to get her to go out of position on lead, so find it hard to know how to adjust behaviour. But I have pounced on your idea of the clicker.

So to try to pull it all together from all your various posts:

I walk in big arcs etc (not the sharp obedience lines); call Jess to heel. She comes into the right position, so I click, and occasionally follow that with a significant treat (probably involving messy fingers rather than neat little schmackos). If she is near to me but not in the correct position I use the NRM - eg. say "Volvo" so she learns when she hears that she is not in the right place.

Hope I've got that all straight in my head; then I just need to get the timing right in the real world...

[Cosmolo]

IMO- treat should always follow your click- but especially to start with, regardless of whether the clicker is charged or not. I don't think you have to walk in big arcs, i would just take a step to begin with.