I'm teaching my dog SwordFighting - Lesson 1 - Wait what?

So yeah. I want to teach my dog how to use a sword.

That's it.

Do I have a reason to teach him to use a sword? No, not really.

Is it practical? Nope.

Is it cool as hell? Yes.

So what I have in mind is pretty straightforward in concept. I want a dog to swing a sword like Zacian using Behemoth Blade in Pokemon Sword and Shield.

For the uninitiated, it looks something like this:

https://www.youtube.com/watch?v=ijyjLzBBexk

I can't really replicate the explosions, and in real life it probably won't look nearly as cool, but hey! Nothing ventured nothing gained!

I'm going to break it down into multiple behaviors and document the journey of us figuring out how to do this. My current blueprint looks like:

1. Pick up the sword

2. Hold the sword (build duration)

3. Swing the sword

4. Hit objects with the sword. (Preferably another sword and not me)

I know this likely won't be perfect, but it's a starting point! So without further ado, let's start attacking this!

Lesson 1: Choosing the Blade

I kinda need my dog to be able to actually hold the sword. The cool thing is I do have a toy sword that was very scientifically upgraded with the highest tech pool noodle attachment to make it easy to grab.

Before Technological Enhancements

After Technological Enhancements

The pool noodle is the spot I want him to grab and hold. I do not want him to hold the sword anywhere else like the blade, so how can I teach this?

For this specific piece I'm going to shape it. The trick is outside of things Missile normally does, and it's pretty likely he has no clue what's going on. I'll just help guide him towards the thing I want!

Oh? What's that you say, dear reader? You have no clue what shaping is? Well let's attack that! The actual definition is:

Shaping is a form of conditioning that leads subjects, often animals who are involved in experiments, to complete an operant behavior. This process is also known as “approximation conditioning.” Why? Psychologists reinforce successive approximations in order to reach the targeted, operant behavior. 

via https://practicalpie.com/shaping-in-psychology/

Okay so that's kinda helpful I guess...

The tl;dr here is that you reward in tiny steps until your dog does the thing you want. So if I wanted to say, get my dog in a box, I could shape it by rewarding them for looking at, walking towards, interacting with, and stepping into the box. The idea is that I know what the end behavior looks like but not necessarily how I want to get there. Luckily my dog has a brain, (I think), so I can let them figure that part out for me!

If you want to see it in action, this is a cool video on the topic:

https://www.youtube.com/watch?v=-p6YuHECpno

So let's do this! As of the time of writing my expectations for what I'll end up rewarding are:

• Missile moving towards the sword

• Missile looking at the sword

• Missile trying to Mouth at the sword

• Missile trying to lift up the sword

• From here it's very possible that he grabs the wrong end, so I'll keep refining him closer and closer to the section I actually do want him to grab.

Here we go!

https://video.wixstatic.com/video/31b2fc_8ed4daa6b3ff4f1f8acafdc3a8f0aa90/1080p/mp4/file.mp4

For this session we just focused on him approaching and interacting with the sword. It actually went... pretty well? I got what I wanted at least.

So let's tackle the next piece, grabbing and lifting the sword.

(Sonic bomb warning)

https://video.wixstatic.com/video/31b2fc_4b3826afb2a447b0b3ef5850f40a69b9/1080p/mp4/file.mp4

Some notes:

• He got really barky during. I'm not totally sure why. Frustrated? Tired? Excited? Something else? (If you have any opinions, share in the comments!

• Despite his barkiness, he recovered and still did pretty well!

• As expected, he did try grabbing the blade. I'll iron that out in a different session before going much further

• He straight-up yeeted that sword multiple times.

Despite all of that, he is doing what I want and actually grabbing the sword. So we'll take it!

And that's it for now! Next time I'll continue building this same trick. If you're interested in more of this kind of thing, I'd love to invite you over to DogKatas where we're building a community of fellow Dog Training nerds practicing all kinds of things!

Look forward to seeing you in the next one, and happy training!