To be very confused about training dogs

Boiled down to its basics, positive reinforcement balanced with negative punishment, works for everything capable of percieving reward.

To date that is: dogs, cats, horses, elephants, rhinos, gorillas, various smaller aps and monkeys, dolphins, seals, sea lions, big cats, giraffes... butterflies even.

What varies is how you apply it, what reinforcer you use, how it is delivered, how you manage the environment you work in, how small your training criteria and how rapidly you broaden them or refine them.

Teaching a dog who likes food not to jump up as someone walks in the room, I may well scatter moderate value food on the floor in the other direction.

Teaching a gorilla to present to the enclosure bars to take a shot, I won't be scattering kibble now will I.

Teaching a butterfly to fly to a mark on a cue... I'll be pairing the presentation of food with the cue, on the mark, and then gradually increasing the distance that mark, that the butterflies are released at.

Teaching a dog who loves balls or tugs more than cheese - I'll be using balls and tugs, not cheese (though I may also use Pre-Macks principle, to increase the value of cheese, because using toys as your only reinforcer is somewhat limiting).

I choose this pairing of the quadrants of learning theory not particularly because they're 'nice' and 'kind' although, properly used, they are.

I choose them because they are effective, and the risk of things going wrong in a way that is hard to reverse, or in a way that causes harm, is incredibly low and for the most part I can figure out what the risky points are and mitigate that risk easily.

The other pair - positive punishment and negative reinforcement - the risks are high and whilst I may be aware of them (I am, but a lot of trainers using these methods are not, and most owners using them are not!) I cannot easily avoid those risks. With these options, you only tend to find out you've done harm after you've done harm, at which point it is very difficult to undo.

For example - choosing the intensity level of a shock collar.

• Start with the collar on low, increase the setting over several sessions until the results are what you want - however this runs the risk of desensitizing the dog to the collar settings, so in an emergency situation when you really need it to work it doesn't - and the time you find that out is... the emergency situation.
• Start with the collar high - its a gamble as to how high that should be. Not high enough, see the first bullet point. Too high, dog experiences a 'single event learning' situation where they pair the extreme pain with whatever they were looking at/thinking about at the time. EG dog was thinking about running in the yard. Dog is now terrified to go into the yard. Dog was looking at the delivery person as some of them bring biscuits - dog now links delivery people with extreme pain. Dog first experiences the high shock on seeing a child with a dog - owners intent is to stop their dog running over to other dogs. Dog actually learns 'small children and dogs = extreme pain, better attack them before they can hurt you'... (nb. All of these are real life examples. The latter two caused the dogs to lose their lives.)

I do understand why people go for the quick fix aversive option - they're desperate for a solution to a problem, and in their desperation either they do not understand the risks they are taking at all or, they have some idea but feel it is justified.

Agree with this. Every dog I have had has needed a different approach and my current one has been…..interesting to say the least. I would never use one of those collars though.

I don’t want a dog that only does what it is told because it is scared, I think that can lead to bigger problems down the line and I wouldn’t want any creature to be scared of me anyway.

Different dogs different techniques needed.
Some dogs are NOT food driven some excessively so.
Some have strong prey drive some zero.

Trainer/behaviour consultant here.

In the UK training is unregulated, anyone can hop on the wagon, hang out their shingle and call themselves whatever they like.

Im retired from training in person now, so I am a non-practicing member of my professional body. I have previously been a member of another professional body - both science based, force free organisations that assess their members before they're allowed to be members, in person, in various situations - I've had to run classes, be observed training totally unknown dog and handler combos, show my history taking process, my report writing, pass written exams...

And then complete a min. number of CPD hours annually, hold the right insurance, uphold a particular ethos and commit to a code of conduct.

But you could just tell some organisations you've trained 500 dogs/for 500 hours, pay your fee (under £200) and call yourself a Master Dog Trainer (theres a chap on TV with exactly this qualification right now, he was a weetabix salesman before that).

Science backs up that positive reinforcement, force free training is the most effective with the lowest chance of risk/doing harm.

That is the bottom line.

The shock collar, prong collar etc folks will argue it til the cows come home but none of the studies they cite stand up, they're all very badly flawed, designed by pro-aversive people, to give the results they want to give.

I will tell you something very telling - PetSafe who make shock collar products, have been sued and are paying out nearly 2 million dollars... because they make false claims that their products are harmless, are safe, are effective - and they have no scientific basis to make such claims and are fully aware there is plenty of science out there showing those claims are false.

We're working on getting dog training to be a regulated industry, and have been for over a decade but it takes time and unfortunately its one of the things that those outside the dog industry do not realise is actually important.

For example - if dog training were regulated, I do not think we would have seen the rise in XL Bully related deaths.

In my opinion (as a professional working in the industry), a huge part of that problem is caused by unqualified trainers and owners, attempting to teach their dogs bite work, using aversive methods.

I think YABU to expect everyone to agree - but if you're looking for a scientific consensus, my understanding it it's all about positive training rather than anything involving aversives or hierarchy. So I'd probably try to work out the red flags for this (e.g. anyone talking about an "alpha" or dominant/submissive behaviour or advocating for the use of aversives - punishment - in training.) and then avoid people who are still sticking to those old myths.

IIRC there is an accreditation that trainers can get which shows they are using evidence based up to date methods - maybe only follow people who have that.

There are many ways to train a dog, and, different methods may give you same the result - think of it as ‘Getting from A to B’. The real difference, is do you have a happy, healthy dog at ‘B’ or do you have one that’s damaged and just submitting.