Unexpected benefits of reward-based training

[Frequency]

After reading through a depressing number of threads where punishment-based training and/or pack theory is being advised I thought this would be a good thread to start.

Big Dog is the first dog I have had since puppyhood and also the first dog who wasn't with me due to existing behaviour issues that need working on. He's also one of the smartest, most enthusiastic dogs I have ever worked with. He pleasure to train. The sheer joy in him when he spots me getting the clicker out is infectious and never fails to cheer me up.

The newest "trick" he has learned is to put his toys back in the toy box on command.

One thing I have noticed with him that previous dogs have never done, is if I am busy at work and don't do his lunchtime play/training session or toss him as many occasional treats as usual throughout the day he will randomly start performing behaviours he has learned through clicker training, including putting his toys away in an effort to earn a reward.

I have now ordered a second box for the kitchen I intend to use as a bin. Our garden is at the end of a cul de sac so in bad weather, a lot of random rubbish gets blown in. I plan to teach him to collect rubbish from the garden to put in the box in the kitchen with the hopes that he starts randomly performing this behaviour.

Does anyone else have any unexpected benefits from clicker training to share (or any ideas on what I can teach my dog so he can start earning his keep by cleaning )

The other thing about clickers is that they make the same sound every time. No matter how much we try not to, our own voices alter depending on mood and other factors and dogs pick up on the variations. Makes for faster training.

@Frequency - my rescue, who I don't think had had any training, started offering all sorts of behaviours to try and get a reward. He had no duration and would carry on by himself as soon as he had received a reward. Once I started lengthening the time between action and reward, he would do all sorts. My favourite was turning around on the spot. No idea where that came from. 9 months on, we are doing really well, and he will stay while I walk across a field and back.

The click is very precise and accurate. Even saying yes or making the click yourself will be longer than a clicker so the accuracy of the marker may be lost.

For me it is the relationship I have with any dogs I train when using reward based training. Along with good old classical conditioning the dogs will see me and know that goods things are about to happen, reward based training makes the dogs feel fab and hence our relationship is happy and the dogs are enthusiastic to learn and be around me.

It makes basic behaviours easier for the dogs to learn I have never taught my dogs a recall, they just prefer to hang our near me and know that when I say their names positive things will always happen so recall is automatic.

Lead walking the dogs are rewarded for being by my side so prefer to stay there than pull ahead for example.

Bed training dogs will happily stay in their beds when the doorbell rings or when chilling out because it has always been rewarded. go to bed and I will get amazing rewards.

Reward based training gives dogs confidence so if they do encounter issues they are unsure of they, have basic behaviours that will be rewarded and help them deal with difficult or new situations.

Lovely thread, taking notes from all your ideas

We make a seasonal trick video for friends each year. The dogs put up Christmas decorations, post Christmas cards, pull crackers, put away the Christmas food shop etc.

Once you have taught a basic hold you can get loads of cleaning and household tasks done by the dogs

If your dogs can put toys in boxes, they can put away the washing, put away the shopping, empty the washing machine and tumble drier, Take off your shoes and socks and put them away etc

My old dog used to know people's names Family & Friends as well as my work colleagues- he used to go to the office a few times a week...
We had a game of post dog by giving him a little shopping basket (kids toy & hold command) put something into it and tell him to go to (insert name). He would go to the person and deliver the goods! Then come back for more!
Apart from the little basket hold, I didn't teach him. He just started doing this when I was folding socks and threw them at the kids to put into their pile. He just kind of picked up the ones the kids didn't catch and brought them back to me at first and when I said it was X socks he went to them!
Disclaimer: this dog didn't like playing fetch!

I have a 15 wk old Labrador puppy, she is so trainable because she is so incredibly greedy. However, I am really rubbish at carrying biscuits and I get very overwhelmed with the feeding and not feeding too much. I’d really like to introduce a clicker. This thread is really interesting . How do I introduce the clicker as a reward? And would a puppy of my pups age really see a click as a reward instead of a biscuit?

@thenewaveragebear1983 The clicker is not a reward. It is a marker that tells the dog a reward is coming.

So dog puts bottom on the floor to sit - click - then reward with a treat. (can be another reinforcer but with a lab I would start with food!)

The clicker accurately marks the behaviour you want and is clear to the dog what behaviour gets the reward

@thenewaveragebear1983 the click isn't the reward. The click is the marker that tells her she's done the right thing and a treat is coming.

You initially 'load' the clicker by clicking and treating several times, so she starts to associate the click with a treat. Then, ask her to do something you know she does reliably, like sit. Click and treat. Repeat a few times. Then you can introduce new skills with it.

X-posted with you @IngGenius

@IngGenius @Lougle (waves from puppy thread)
thank you- do you eventually phase out the food reward though and just ‘click’?

@thenewaveragebear1983 no, the clicker always has to have a treat. To the point that if you accidentally click, you still treat. If you get it wrong and click early, you still treat. Never click and don't treat. But what you can do, is fade out clicking once a behaviour is fluent and paired with a command. You can also change the criteria that you click for. E.g. you click for being on your left side when lead walking. Then once that it going well, you can only click if all four paws are on the floor, etc.

Ah I see. I think I need to do some more research before I start.

Often the clicker is used to teach a new behaviour so in the example Lougle gave you would not always have to use the clicker to get your dog to walk on your left side.

Once the behaviour is on cue you would not need to use the clicker but reinforcement by reward is always a good way to go.