A location specific marker (LSM) is something worth considering. A LSM tells the dog where to expect (or report to) to collect reinforcement. It may also include information about what type of reinforcement will be offered (usually whether toy or food). Anyone who’s played around with different ways to reinforce their animal knows how helpful it is to formalize when the dog is invited to chase after a thrown reward. Because, if you don’t, your dog will think it’s fair game to chase after any attractive thing that happens by. So, I’ve always encouraged students to put it on cue (“Get it”), or, at the very least, release your dog (“OK”) to clarify that they may break out of position and/or handler focus to pursue the reward.
Another useful function of LSMs is as a barometer to gauge your dog’s state. Markers, by definition, predict reinforcement. By reading how your dog responds when you mark (via “marker games” with nothing much on the line), you can gain valuable information about where your dog is at emotionally and whether you need to adjust and raise (or calm) your dog’s arousal level, or otherwise make the task easier. Marker games allow us to check out the dog’s mental state without asking them for anything too difficult. And, since the mark will be reinforced, it’s a reinforcing technique to check in with.
Some people create a very long list of different LSMs. While, I certainly see the benefit of an LSM, I don’t think it’s practical to have too many. For one thing (Chaser the BC aside), I believe that trained cues ought to be thoughtfully and sparingly crafted. Dogs aren’t verbally inclined. So, when I am deciding on a new cue (and I train the majority of behaviors to a verbal rather than just contextual and visuals/body language), I think long and hard about what word I will choose and whether or not I already have something similar enough on cue that would eliminate the need for adding a new cue. For example, for heel positions (remember, I train heel in multiple locations: front, left, right, behind, etc), I name each heel position. I do not have an additional set of cues for the dog to swing their butt in, pivot, change pace, or stay in heel when I back up/move sideways/etc. Just 1 word: “Heel”. Having too many LSMs would break my rule of using as few words as is necessary.
I’m all about clarity. And while I think that too many words (including LSMs) can get muddy, some would argue that numerous LSMs create more clarity. Perhaps. But, back to the practicality point. Most people can’t remember a single mark while simultaneously running agility. I repeatedly remind my students to mark before providing their reward so that they clearly tell the dog to stop agility and collect the prize. If instead they sloppily hand the dog a reward, the dog sometimes confuses a hand signal to take the equipment with being handed a treat. It’s also common for people to mix up praise and their marker. Sometimes they praise and then treat. People regularly mix things up with only 1 marker word to keep track of!
When I first heard about LSMs I didn’t realize that I was already using cues to serve this purpose. I was looking for ways to reduce any frustration my learner might have. I was struggling with heel training (because I use a toy too much and too frequently as a reward, but that’s another post) so the idea of using a different mark to inform my dog that she was going to get food instead of her cherished toy made a lot of sense to me. Before we go any further, let’s look at an example that might help you understand what my dog felt like when she was offered food instead of the toy.
When you agree to do a job, it’s for remuneration, right? Imagine if you were paid, but in the equivalent value of gift cards to Amazon. How would you feel on payday with your rent or mortgage about due? Probably anxious and not happy to do more work, and likely totally distracted with thoughts about how to acquire money for your bills. You certainly aren’t ready to put in your best work! To take this example a little farther, what if you expected to be handed a check at your office, but you find out you have to go to the 3rd floor to pick up your check. Not too big of a deal, I guess, so long as you always go to the 3rd floor every 2 weeks. But, if you show up on the 3rd floor next time and they send you to the basement to collect your check, this inconsistency becomes frustrating.
So, the concept behind LSMs is to alleviate some frustration or disappointment that might result from not knowing where or how to get the reward or expecting something different for pay. There’s a lot of merit to LSMs along with intelligent reward placement. If your dog correctly anticipates what she is getting and where, you should have a happier and more confident learner.
So, which LSMs make the most sense? I think context and expectation are already woven into how we behave when we train. That said, it is important to examine our behavior during training (thank you video!!) so you know what your dog expects. It’s important to see how predictably you do certain things with the reward delivery…which, by the way, you might want to change in order to spice things up and make training more interesting. That sort of awareness takes constant reflection. So, it stands to reason that most of us may be better off with some more LSMs to keep things clear.
If my click means food most of the time, I can use “Catch” to clarify that I’ll toss a toy. Once in a while I’ll say “Catch” and toss food instead. I usually let her see that I have food and then say “Catch” to avoid a miscommunication, but another LSM to clarify “catch a piece of food” might be in order. Other times I might click and then give the toy. I think there’s value in clicking and then offering a toy instead of the expected food to increase her interest in an exercise. BUT, she may be disappointed if she expects a toy and gets a treat instead. I’m still debating about where and how the click and LSMs work best.
I’ve settled on just a few LSMs. “Catch” for when something will be thrown that the dog should snatch out of the air. This type of toss is different than chase. I may toss to my dog while in position (like in a down at a distance, in heel, etc), or I may show it to her, say “Catch”, and then lob it for a dog to leap up and catch. If I want my dog to drive to the toy and/ strike it, I have a different cue, “Get it!”. I would use “Get it!” to get a dog to drive up to a piece of food in my hand, too. “Get it!” is also my send to reward cue (that item may be still or in motion). Some people create a different cue for a send to the toy vs strike the toy in hand. In my mind, “Get it!” means permission to zero in on and grab (with a retrieve built in). I don’t see any confusion. Separately, I use the cue “Find it!”, which is my catchall for anything search related, from finding the hidden item to permission to snuffle up crumbs and other such behavior.
Another LSM worth using might be one that informs your dog to hold position. When training your dog to work at a distance, you don’t want them to travel forward out of position to collect their reward and create a conflict for the dog as to why to remain at that distance. A LSM (like “Hand” or “Room service”) that means ‘wait right there, I’m going to deliver the treat to your mouth’ might help. That said, I don’t use one. My hope is that the dog will stay put unless I clarify to “Go, get it” when the dog ought to move to collect the reinforcer. In situations where the dog is to remain at a distance, I’ll try to praise and bring the reinforcer in to the dog (praise, unlike a mark, does not “end the behavior”). If I toss, I try to toss at or past and behind the dog (and say “Catch!” to allow them to break position to catch). If I set a reinforcer out behind the dog, I’ll “mark” the behavior with the cue to spin (to turn towards the item) and interrupt the spin with a “Get it!” cue. I always want to clearly release my dog in any situation where I’d otherwise like the dog to continue working (at a distance, holding a stay), and a LSM or another cue or a release cue can each serve that purpose.
Different tricks require different reward strategies. If I click a dog for heel, the reward will be offered in heel. The dog is free to dance around after the click. But, the reward will arrive where her mouth was when clicked. I’ve heard Michele Pouliot say “feed the ghost”. The idea is that the dog will learn that dancing out of position is wasted effort. I could opt to use a “Room service” LSM to teach the dog to wait in position. Or, I can just train the exercise and let the dog learn by the process…or not. The dog IS allowed the freedom to move out of position according to the “laws of the clicker” (I used to break this rule but it resulted in frustration). My remedy in such a situation is to feed food faster (before the dog gets out of position), sometimes opting to hold multiple pieces at once (usually in the far side hand) so that treats can be rapidly fired at the dog before the dog dances around too much. Once the dog has value for holding the position, I space out the rate of reinforcement. I also practice slow-motion treat deliveries using withholding if the dog tries to rush the treat (not to be used with a click, though). This helps a dog learn that the way to earn is to stay put (though this is not positive reinforcement, I find it extremely practical and effective). I may change my mind and train a “Room Service” LSM. There is a limit to how many things I can focus on at once, and my wishlist of things to train is long and without a greater need, “Room Service” is not a priority.
I start debating when I think about where LSMs and the clicker intersect. There’s a lot of “set up” to my clicking sessions when I’m working on just 1 behavior. I often click and toss a reward to reset the dog for the next rep. Sometimes when I click, the dog can expect to be handed the reward in position (like in heel). In other situations, the location of the reward is obvious, like a bowl or 2. If the dog didn’t realize the treat was in the bowl, I’d rattle in the dish, and use my generic “Get it!” (possibly with some movement/body language) to set the expectation and get the dog to anticipate collecting reinforcement from the bowl. That said, I’ve been considering developing a lower intensity LSM (my version of “Get it!” is energizing) for when I want to work on things like cavalettis where it’s best to calmly advance to the bowl…
Do you think people are going to invent different click sounds in order to use LSMs with a clicker-type device? Like a clicker with different buttons (trademark!). That seems only slightly more complicated than using a lot of LSMs. Personally, if I click but plan on rewarding something other than directly in the mouth or an obvious location, I will follow my click with information about how and where to collect reinforcement. I MAY mark with my LSM directly, but, sometimes it’s not until after I click that I decide to offer something other than feeding where my dog expects it. You could argue that I’m talking about a cue instead of a mark. I could click and toss in the bowl, or I could say “Get it” and toss in the bowl. “Get it” has some intensity built into it so I’d prefer to click if I wanted the dog to be more thoughtful about the exercise rather than the reward. I blur the lines between mark and cue in relation to what I’m referring to as a LSM. Hopefully that’s not a problem? In my mind they both fall into the category of ‘words used in training”.
Like anything, I think this can go a little too far. A big part of dog training is timing. If you are trying to decide between 12 different markers, I’m pretty sure you’re going to have some trouble spitting them out correctly at the right time. Plus, it seems like too many things to train. Depending on your training goals, you can probably dial it down to a few that will work…or, you can be predictable (though that has its own problems) with your rewards. I also have a ton of other things to train, so, I want to keep this as simple as possible.
Here’s a list of some of the LSMs I’ve heard of…and some people have another equally long list to differentiate food vs a toy in each of those locations:
- “Get it!”
- “Find it”
- “Bowl” when you put the reward in a bowl (I use “Get it”)
- “Face” for throwing a toy to the dog
- “Switch” to change from one toy to another
- “Behind” to get a toy behind you
- “Scatter” to snuffle up a pile of food
- come get the item “Snack” for food and “Strike” for the toy
- “Room Service” meaning stay put and I’ll put the treat on your tongue
Probably the best tactic is to examine what reward strategies are most important for your particular interests and zero in on including some LSMs that help you keep things really clear in those important skills rather than having a LSM for any and every purpose. What do you think?