- It’s the easiest thing in the world if you have a shaping-savvy dog.
- I don’t have a clue where to start if you don’t.
Pretty much all of Lilo’s intentional training has been clicker-based positive-reinforcement training. I’m only human, right? I’m not above an occasional collar pop or a human yowl of frustration and/or behavior-interrupting. I’m basically agnostic re: what people do with their own dogs; if human and pooch are happy, it’s all good in my book. But for this dog, this was and is the approach, and I’ve been really happy with it. I can talk a bit about the reasons why elsepost, if anyone’s interested. I’m only mentioning it here because it really informs the way I taught this behavior. So this post is going to be more of a ramble than a tutorial and I apologize in advance if it doesn’t generalize well.
So: go-to-mat. This was one of the first things Lilo learned in a really wonderful puppy obedience class and the behavior that, the instructor said, if universally taught, would keep more dogs out of the shelter than any other. I believe it. Being able to reliably send the dog to a fixed spot and get them out of the way and quiet/settled down is so useful. It’s functionally similar to crate-training, I s’pose, though more portable (and obviously less failsafe to start, although Lilo’s history with crates is a whole ‘nother story).
The finished behavior: a dog that, when asked (explicitly or through context cues, like scrambling eggs), seeks out a designated spot and lays down there and stays put until released. In theory, you can go as far as you want with teaching a “relax” or “settle.” Lilo’s squishing her face onto the floor was part of that process…although she’s a worrier and a really hard trier, and still tends to give me the some pretty intense “relaxation” if she thinks we might be doing something instead of just hanging out for a while. (Teaching her to roll over onto a hip (instead of forever holding her default sphinx down) was a step in the right direction.)
We started with a specific mat (the one pictured), but this behavior, once established, does seem to generalize well. I’ve borrowed towels at friends’ houses and used saddlepads at the barn, etc. Two reps is usually enough for her to figure out that, “Place!” now means, “Go to that thing instead.” And sometimes she generalizes it herself; she likes the location of the kitchen rugs and keeps trying to convince me that it’s her mat now.
The training process: start with a dog that understands the rudiments of free-shaping*. That is, one that groks 1) the marker you’re using (I like the precision of a clicker when shaping, though I’ll use a verbal marker with established behaviors) and 2) the idea that trying stuff out pays. Put the mat down in your training area. Wait. Mark/reward any interaction with the mat: feet on, sniffling, even looking at/acknowledging if you have a reticent dog: whatever. Repeat. Repeat. Repeat.
When the lightbulb goes on — that is, when the dog starts to figure out how to make you reward them — start raising criteria. Maybe look for two feet on the mat, and then three. Look for the dog to become magnetized to the mat — to clearly understand that moving onto it is the point of the exercise. With Lilo, I rewarded by tossing her treat a short distance from the mat at this stage, thus setting her up for another rep.
Once you have four feet, start looking for the down. You might hit some frustration at this point; typing it out, it feels like skipping a step, but I can’t figure out what the interim step would be. I’d probably accept a sit, but I wouldn’t want to linger there for too many reps and have the dog decide that the sit is important. (I’m not actually sure if Lilo had a sit when we started mat work. I know she had a nose-touch and that she did not have a cued down; I built that later out of a play bow.) If and when the dog happens on a down-on-mat — at this point or earlier — I’d jackpot the hell out of it (and then, if earlier, go back to lower-level rewards for feet on until the dog is magnetized). The criteria are chest and both elbows on the mat. I personally don’t care if the hind end hangs off, but chest-and-elbows cuts down on creeping.
I definitely didn’t name the behavior (that is, add the cue — simultaneous to the behavior, not before) until Lilo was consistently offering to run to her mat and throw herself down on it. I can’t remember how much distance we had before naming it, but I think probably 5-6′. This matters; you want the dog to be offering the behavior that you want before you tell them what it’s called. If you don’t want to keep, say, three-feet-on-standing-up, then don’t dwell there long enough to give it a name. I did name it before working towards the head-down hip-rolled position, but we dropped the verbal cue and went back to shaping when altering the criteria.
(I am super sloppy about my transitions between behaviors in a given training session. With Lilo, I get away with it. I suspect the next dog will tell me whether that’s because it’s actually not such a sin if the dog is used to how you work orrrr if Lilo is just a genius in that particular realm.)
And then you built distance — how far away you can start when sending the dog to the mat — and duration — how long they stay in position before you mark/reward — from there. When building duration, I switched over to rewarding her in place (handing her the treat while she stays down and/or placing it between her front paws). Ditto that while working on relaxation, actually, to keep everything lower-key.
It may go without saying, but just in case not: you do not have to do all of this in one sitting! You probably shouldn’t, even if you can. What works best for a given trainer and dog depends on them, but lots of bite-sized sessions are generally a great way to go. Lilo is a hard worker who loves training; she’s also a methodical thinky dog who benefits from — and will give herself as needed — breaks to let what she’s working on sink in a bit.
There will, again depending on you and your dog and the phase of the moon, etc., almost certainly be moments of frustration and just-not-getting-it. What that looks like will vary from dog to dog. I’m not (as this post probably makes clear!) an expert trainer; I’m just an interested amateur.
But as an interested amateur who spent a lot of time in dog classes? The biggest thing I saw people struggle with re: shaping was waiting out the moments when the dog didn’t seem to get it. And it’s very, very possible that my personal biggest strength as a trainer is my knack for standing there smiling at a dog who is trying everything except the thing I’m looking for. I find it charming and hilarious. I really, really recommend finding it charming and hilarious, you guys. Try not to get mad; try not to get frustrated; just give them space and time to work it out and if it just isn’t happening, back up to the previous step/criteria and shore that up before raising criteria again. Trust the process. Trust the dog.
Putting it to work: once you have the behavior reliably on cue, you can start putting it to work. Lilo went to her mat for pretty much everything in those early days: to wait for her meals, while the humans in the household were eating theirs, while I put on my shoes and gathered up gear for dog-walking, and so on. Definitely when we went over to other peoples’ houses, to the ice cream stand, and all sorts of out in public: the mat went, too, and sometimes the entire outing would revolve around having her chill on her mat with intermittent rewards.
(I’m quite sure that you could end up with an equivalent behavior by taking a dog with an established down-stay and consistently putting it in one on top of the mat in your desired context. This is basically how the Best Dog learned to lay under the kitchen table waiting for carrot scraps. So if you want to skip the whole shaping process, there you go. But I do feel like it’s easier with a dog that’s used to a system that rewards creativity.)
Dogs tend to like routines and shaping-savvy dogs tend to anticipate and offer what they think will pay well. Tell a dog to go to her place enough times before you put her dinner bowl down eventually results in a dog who runs for her place when you pick the bowl up in the first place. Or (as Lilo taught me a few weeks ago) when she thinks she might be able to sell you on the idea of picking it up. Tell a dog to go to her place enough times while you’re cooking your own dinner (and reward her with a couple of scraps as you do) eventually results in a dog who runs for her place when you break out the frying pan or chef’s knife.
Or when you pick up your shoes, or the doorbell rings, or the cat walks into the room, or whatever. Reinforced behavior continues, right? So you train the behavior. You put it into practice. And you reinforce, et voila. It’s very simple (which is not necessarily to say easy).
I have no idea if this post is going to make any sense at all outside of my head, especially since I’m obviously skipping the steps in between “acquire dog” and “decide to mat-train dog who has mysteriously become clicker-savvy.” This is one part because there are already plenty of words here and one part because I don’t clearly remember how we managed that; I think I know but I have a little reading up to do before next!dog enters my life! But there are lots of good books out there, for folks who are interested. I’m particularly fond of Sue Ailsby’s Training Levels, which are nice and clear and do a bomber job of explaining what thin-slicing and raising criteria are all about.
And that’s (more than) a few words about mat training! I hope y’all find this useful or at least interesting. Less jargon next time, I promise!
*One could certainly teach this via luring and/or capturing instead. I do use some capturing (mostly an unstructured “catch dog doing something right” kinda way) although I’m too impatient to rely on it exclusively. I’m not a huge fan of luring; I really like watching a critter figure things out for themselves; I just think shaping is way more fun and default to that instead. So I’m not actually sure what to recommend re: starting to lure a go-to-mat. But the Training Levels book definitely talks about it.