Food Rewards

[Guest Tess32]

Any instructor worth their money should simply tell you to work out what your DOG finds motivating.

[dogdogdog]

My dogs are not food orientated.

[smokeysue]

Still wondering what do you give your dogs as a reward. Smackoos, meat, cheese, etc. Would love to know. Sue

[J...]

I have to admit I was a bit against food training to start with, but was getting great results so took it on. I like the poker machine analogy - keep em guessing!!

Smokeysue I use Supercoat Lix dried liver (from the supermarket) for two reasons, 1) my dogs a chunker who gets fat on the smell of an oily rag so the dried liver is probably as good as I'm going to get fat wise, and 2) because I can chuck them in my pocket without the plastic bag and not have a problem trying to find treats in amongst the plastic/treat bag etc. I'm a bit unco at the best of times so anything that makes it easier for me..... Found two pieces in the clothes dryer the other day.... must remember to check my pockets before I put them in the wash

Many recommend kabanna, cheese, frankfurt etc (too fattening for Darcy plus it'd make a mess in my pocket :D ). The only issue with the dried liver is if she gets a bit too excited it gets stuck half way down her throat and she'll cough and splutter for a second, but she doesn't do it often! I tried smackos but Darcy wasn't as eager for it as she was for the liver, plus not a fan of the high salt factor! :D

[Mooper]

Thank God the days of jerk, yank and yell are coming to an end. Food, or let's say motivation, is currently the best way to teach a dog what we want from him/her.

Welcome to DOL, Canid. :D

I've singled out part of your comments because I've just spent a weekend at sheepdog school with my dog, and the experience is therefore fresh in my mind.

My dog is very food-motivated, and responds very well to clicker training. In obedience classes, we went through a fair few treats early on -- btw smokey sue, we tended to use something smelly like cabanossi, cubes of cheese or cold chicken in class but used Nature's Gift liver treats or dehydrated liver for daily practice. Jack would get less and less food rewards as my standards for his behaviour increased. As others have mentioned with their own dogs, I would only click and treat for perfect behaviours thereafter.

But sheep herding is a whole new ballgame. There is no need for food rewards. The reward is being allowed to work the sheep, and it's a mighty powerful one once a dog is switched on to herding! Which makes me wonder what other rewards (flyball? agility? tracking?) are so powerful that food is no longer required as a motivation.

Yesterday I noticed one dog in particular, whose owner was feeding him treats like a freight train. I also noticed that he was getting treats even when he hadn't done exactly as commanded ... for instance, he was told to sit at heel, and would sit on the owner's feet with his back to her ... it was very interesting to watch.

Today, in the pen with sheep, there were no treats. The dog didn't really work (but is pretty new to sheep, to be fair). Of more concern was the fact this big dog paid absolutely no attention to his handler's commands. He started out looking at her hands, nudging her pockets and sniffing the ground for any fallen treats. It was a lost cause, I believe, because the owner has mistakenly believed she was training her dog all this time with food rewards when he was really training her.

I'm pleased to say that the dog and his handler came a very long way today. After a loooong time in the ring and lots of assistance form the trainers, the dog *finally* looked directly into his owners eyes and actually came to her when called rather then going straight past. And he sat at heel rather than on her feet.

After all that blurb, my point is that I'm now learning that while food, tug toys, pats, praise or whatever can work really well in the obedience ring (ie different motivators for different dogs), there doesn't seem to be any place for food rewards in herding as far as I can see.

The best part of all was seeing a sheepdog head out into the paddock first thing this morning. This young bitch was simply told 'get away' and she raced away hundreds of metres from us, up into the paddock, rounded up the sheep and brought them right back to her master. All for a sedate pat on the head and a subdued 'good girl'.

True sheepdogs have me in awe

[lab and poodle]

"Which makes me wonder what other rewards (flyball? agility? tracking?) are so powerful that food is no longer required as a motivation"

I have to admit that my old dog would track for the sake of tracking(he go through to TQT on picking up an old teddy bear) , but his motivation and ability to keep to the track were sure as hell extended by the application of tins of sardines at the end of the track. I believe that thius made it fully operant, do track (in prey drive) get reward. Don't do track or get diverted, then no reward. Up to him.

[lab and poodle]

What rewards

For my Lab I have used Normal dog biscuits, sardines, luncheon, liver, breakfast, roast beef, chicken,apple, Dogs to play with, New people to greet, a game, some more attention, a few throws of a tennis ball , unbelivable repititons of retrieving in a pond.blah blah.

For my Poodle, Liver, stras, frankfurters,chichen(no go), silly bloody grocery treats (spat out),cuddles,TOYS, people,dogs, always looking.

Look out for fast treats (dog eats quickly) like stras. Avoid slow treats (like most grocery types), dried liver etc. It sort of ruins the flow.

I am always concerned that people base their training philosophies on watching hardened sheep dogs work, or maniac retrievers go for it. They are a very small percentage of dogs doing precisely what they are bred for. A fair amount of "shot gun" selection has gone into sheep dogs too! Unsuitable dogs get shot. Our mostly pet dogs are been asked to do tasks that they were not specifically bred for, in a pretty demanding enviroment. A few treats to help them learn artifical tasks is no big deal to me.

[J...]

You're exactly right Mooper..... I've grown up with working dogs, and I don't ever recall any of them being given a food reward. They're bred to work, and they love it!

My dad has a bitch that, if there were sheep around the house, would go out and round them up of her own accord. You let the 4 dogs off for a run, give it about 5 mins of "unsupervision" and you would end up with a mob of sheep at the gate of the yards behind the house. Thats what she loves to do! :D

I guess with food training with agility, obedience etc.... when you work a dog with sheep or cattle then you don't really ask a lot - i.e bring me the mob of sheep/cattle, slowly without stirring them up, and don't leave any behind. The rest is instinct! We don't make demands on which rock the dog should jump over, or how long it should sit for at that thistle, or expect it to weave in and out of the fence posts while its at it! Yes you expect a little more of them when you take it to sheep trial level, but being able to do what they are bred to do is reward enough for most dogs! Placing a few extra demands on them isn't a big deal!

OTOH - to compare to agility as an example, we expect them to do exactly as we want, nearly to the footstep and second. Wait here, make you step on the end of that ramp, make sure you go and out the right side of the weave poles etc. Plus dogs weren't bred to jump through tyres, run through tunnels or along balance ramps, and its not instinct for them to do so :D

BTW not picking on agility here at all, just used it as an example

[MrsD]

Today, in the pen with sheep, there were no treats. The dog didn't really work (but is pretty new to sheep, to be fair). Of more concern was the fact this big dog paid absolutely no attention to his handler's commands. He started out looking at her hands, nudging her pockets and sniffing the ground for any fallen treats. It was a lost cause, I believe, because the owner has mistakenly believed she was training her dog all this time with food rewards when he was really training her.

Mooper, I think this comment is the exact answer to someone's question above as to why you cant use food in the ring.

The idea of the competition is obedience - not the dog getting a treat for doing something because he's taught the owner to give it to him :D .

[lab and poodle]

"Mooper, I think this comment is the exact answer to someone's question above as to why you cant use food in the ring."

I totally agree that food shouldn't be used in the trial ring, but I really dole it out for correct good behaviour while I am training them. I wish I could attach a video or something and you might get some idea of what you can get. I reckon I would cop a bit of flak from all sides, (Being too soft blah blah on one side, too vocal using change of pitch for the purist clicker trainers, occasionally using vocal corrections, "making" a 12 1/2 dog do obedience (he should be by the fire getting fat), training a puppy under 6 months, it never stops)

I think to be fair though, all training aids should get the biff in the trial ring and that would include chain collars. I can hear the justifications and bleats now, so I will reply in 20 postings..

[Kavik]

With the different disciplines:

I think some things the dog can find inherently rewarding (sheep work for example) but with other things such as obedience which is artificial, the dog needs external motivation. This is where I find food/tug etc come in very handy. In agility, many dogs find the obstacles themselves motivating, many especially like the tunnel and A-Frame.

I would say there is a lot to sheep work! It is not easy, you are asking a lot of the dog, some things you want the dog to work out for itself as well. Not to mention the different styles of working sheep and different types of sheep dog trials.

[MrsD]

"Mooper, I think this comment is the exact answer to someone's question above as to why you cant use food in the ring."

I totally agree that food shouldn't be used in the trial ring, but I really dole it out for correct good behaviour while I am training them. I wish I could attach a video or something and you might get some idea of what you can get. I reckon I would cop a bit of flak from all sides, (Being too soft blah blah on one side, too vocal using change of pitch for the purist clicker trainers, occasionally using vocal corrections, "making" a 12 1/2 dog do obedience (he should be by the fire getting fat), training a puppy under 6 months, it never stops)

I think to be fair though, all training aids should get the biff in the trial ring and that would include chain collars. I can hear the justifications and bleats now, so I will reply in 20 postings..

Lab & Poodle, by chain collars, do you mean a check chain?

When I train, I start off by using food as a reward but very quickly change to something else (usually praise). After training the dog for a length of time, its sort of like someone else mentioned above - the exercise & the "buzz" from doing it correctly is self rewarding for the dog. This has been my personal experience, but then I have trained working dogs - it just swaps one set of "work" for another. It may not work very well (pardon the pun ) for other groups of dogs, I dont know.

My dogs have always got excellent marks in the trial ring for exercises & were known for their enthusiasm & I think its because it was "work". They didnt really care what "work" it was, as long as it was work! :D

BTW, my dogs were obedient no matter if they had a check chain or any other type of collar on or not .

[Lablover]

Reward is any consequence any animal perceives as a consequence of its behaviour.

If a dog breaks a command when offered a treat it means the dog percieves the treat is a 'punishment' compared to the 'reward' of freedom maintained by breaking a command.

Once that happens it has 'positivly reinforced' ignoring behaviour.

In other words the 'consequence of the behaviour' of ignoring commands makes ignoring commands reward 'behaviour' -- 'punishment' and 'reward'.

Denis,

I am going to use the excuse , of just returning to the house, after spinning. I do not understand your reasoning.

Could you please expand on the above quote?

[MrsD]

In agility, many dogs find the obstacles themselves motivating, many especially like the tunnel and A-Frame.

Kavik (sorry, I didnt see your comment when I posted my last post), that's exactly what I meant, but with my previous dogs I found that the exercises in obedience (particularly once you get to open & UD) themselves to be a big motivator for my dogs. Agility is particularly motivating as you mentioned, but I also found recalls, dumbell & scent work & jumps to be a big plus in obedience itself

[lab and poodle]

Lab & Poodle, by chain collars, do you mean a check chain?

Yes, any training type collar.

[Denis Carthy]

Denis

Reward is any consequence any animal perceives as a consequence of its behaviour.

If a dog breaks a command when offered a treat it means the dog perceives the treat is a 'punishment' compared to the 'reward' of freedom maintained by breaking a command.

Once that happens it has 'positively reinforced' ignoring behaviour.

In other words the 'consequence of the behaviour' of ignoring commands makes ignoring commands reward 'behaviour' -- 'punishment' and 'reward'.

Lablover

Denis,

I am going to use the excuse, of just returning to the house, after spinning. I do not understand your reasoning.

Could you please expand on the above quote?

Denis

Hi – Ugg – the way the words - ‘positive’ - ‘negative’ – ‘reward’ – ‘punishment’ as used in learning are from a theory known as ‘operant theory’

– in dog training they are the most misused words any language has to offer – its notoriously difficult to explain or teach because of the way the words are normally used e.g. ‘positive’ is always good, ‘reward’ is always nice, ‘negative’ anything is always bad, ‘punishment’ is retribution.

– It attempts to give a language to any animals response to any external event/stimulus at that nano second in time, if it is interpreted in simple terms it means - ‘anything, any animal learns as a consequence of any behavioural response to any external stimuli’ -

–

Below is the TOTAL meaning of the words as used in dog or any other animal training/learning, they exclude the animals psychology, physiology and environment beyond the nano second it time of a behavioural occurrence.

1.

Negative means ‘something’ is removed/withheld.

Positive means ‘something’ is added.

Extinction is said to occur when no negative or positive stimuli are present.

|

Negative Reinforcement = increases a behaviour (strengthens it)

Negative Punishment = decreases a behaviour

Positive Punishment = decreases a behaviour

Positive Reinforcement = increases a behaviour

Reward means:

‘any appetitive/favourable/nice (to that individual animal) consequence any animal perceives is the consequence of a behaviour’.

E.G’s.

A murderer knifes someone to death – he does not get caught, he is rewarded by the consequence of the act of murder, it might increase the behaviour.

Two fighting dogs are fighting to the death, one dog kills the other, the behaviour of both dogs is ‘reward behaviour’ (they like fighting to the death, they will not run away).

Punishment means:

‘any non appetitive/favourable/nice (to that individual animal) consequence any animal perceives is the consequence of a behaviour’.

E.G.’s

A murderer is about to knife someone to death – he sees a policeman just as he is about to stab someone, he is punished by being unable to carry out the act because he fears he will get caught.

= Negative punishment – the reward is withheld via fear of capture.

Two fighting dogs are fighting to the death, both dogs are separated by their owners, they still try to carry on fighting, they are punished because the owners have removed the ability to fight by holding them apart.

= Negative punishment – the reward of killing is withheld via physical restraint.

Sorry I dont have time right now to go into it further Lablove hope thatgives some legible explanation. To add, outside laboratory conditions it cannot be controled with a liveing animal,

e.g. it excludes the psychology for one thing, food rewards are predominantly prey drive stimuli, pack drive stimuli are the mainstay of the relationship & obedience, you cannot take a dog to mating and expect him to tie because you offer him a treat to do so.

Edited October 3, 2007 by Denis Carthy

[Lablover]

Thankyou Denis.

I appreciate your time.

[lab and poodle]

e.g. it excludes the psychology for one thing, food rewards are predominantly prey drive stimuli, pack drive stimuli are the mainstay of the relationship & obedience, you cannot take a dog to mating and expect him to tie because you offer him a treat to do so.

I use both "MODELS" to Drive and operant, in understanding my dogs behaviour when I am training. They aren't neccesarily disjoint. The point of most experiments in operant behaviours is to exclude outside stimuli other than those in the experiment . This is done by using asethetically uniform and bland enviroments. Some clicker trainers (I soemtimes use a clicker too) try and duplicate this by using no emotion during training, and some even use well designed training rooms.

As I have progressed in my training, I have been more reliant on observing what my dog is doing, and adjusting my actions to maximise my training efficency. Well to be real I am doing the whole dam thing to enjoy myself! If i don't get it right, my dog might drop a dumbell or similar, and for that I can't see a breakdown happening. I also use the drive model because it explains a dog's behaviour relative to it's enviroment, which was what was missing in operant modelling.

I think it is important to use this huge body of work involved in operant behaviour. Look it up, read the base texts (I did before I changed methods) . Then you don't have to work out again (it's called re-inventing wheels, a waste of time) what various different actions might or might not do.Be aware though that not everything in a scientific text is right. I am in a much more black and white field and in an 8 page paper, I could find up to 8 "errors".

It is interesting to note that the "drive" model relies on observations of animal behaviour and Human interpratations of them, (similar to social science type methods), and operant behavior studies on a very clinical scientific approach. This drive model with today's pollies would have had it's funds cut off for being too "trendy" and wooly.

[Denis Carthy]

Lab and Poodle

my dog might drop a dumbbell or similar, and for that I can't see a breakdown happening.

Denis

Just to take that one example in operant –

At the point the dog is carrying the dumbbell, the ‘behaviour’ of carrying is a reward to the dog.

- at a point it drops it the behaviour of carrying has become a punishment to the dog.

- dropping the dumbbell is then the reward ‘behaviour’.

- At the point of watching the dog carrying the dumbbell watching is a reward to the owner.

- At the point the owner watches the dog drop the dumbbell, watching the drop is a punishment to the owner.

Reward and Punishment are in fact 'functions of behaviour' which give the appetitive/favourable/desired consequence.

Also, the value/intesity of static/fixed external rewards and punishements stimuli are not constant which is why one second something is a reward to the dog and the next a punishment, it is the change in behaviour which determines the 'value/intensity' of an external reward in relation to the total environment.

Edited June 6, 2006 by Denis Carthy

[lab and poodle]

Denis

I am not disputing what you are saying, I am very well aware of what the theory says and means. I was putting it into personal context. I should have said "I won't have a breakdown if my dog drops a dumbell. It just isn't that important to me" . It was a semi snide go at trainers who seem to loose the plot at what they are doing and for what. It is an area where operant theory has some difficulty, namely addictive or reflex behaviours.(I must get my dog to do the retrieve at any cost, I must , i must ,I must) It describes them very well, but hasn't really come up with a solution.