Dog Training Flashcards
Training Plan
Training roadmap—series of steps planned in advance to teach a behavior.
Criteria decisions already made—you should seldom have a “what” problem.
If it’s not fully memorized, it should be in writing next to you during training sessions for reference.
Building a
Training Plan
aka Backwards Planning (OC)
- Describe the terminal behavior—the final step in raising criteria
- Break the terminal behavior into all parameters
- Determine first step—closest approximation the dog can do now
- Use the parameters to create incremental steps—aim for quick and reliable progress
- Train, adding splits as needed
More parameters create a more detailed plan and longer training time.
Record Progress
Always note which step you reached on each behavior after every session.
No fishing around or warming up a behavior next session—more efficient.
Standard Operating Procedures
SOPs
Well vetted, standardized training plans for all popular behaviors. Includes Plan B (or even C) for commonly needed splits.
Memorized over time, making for much faster training.
Acronym, define, and function
PDS
Push Drop Stick
Sound, systematic adjustments to the set criteria based on the last sample/set.
Careful adherence prevents decision-making or guessing while training.
Alternate PDS
Can use another count (i.e. sets of 7 or 10), but be sure it’s both reasonable and systematic.
Push-Drop-Stick balance
Good training has a healthy mix of all three.
- Constant dropping—plan is not incremental enough
- No dropping, splitting, or even sticking—increments are too small and inefficient
Good training is not constant pushes.
Push
Go to the next step in the plan—raise criteria
* 4/5
* 5/5
Drop
Return to previous step in the plan—drop criteria
* 0/3
* 1/4
* 2/5
Auto-drop to avoid crashing RoR and losing the dog.
Auto-drop
Auto-drop immediately—don’t finish the set
* 0/3
* 1/4
Drop to fix crashed RoR.
Stick
Repeat current step—maintain criteria for another full set
* 3/5
Trial
or
Rep
One repetition of a step.
Set
A series of 5+ standardized reps of a step.
Same antecedent each trial.
i.e. exact same hand signal motion, speed, and height throughout set.
Usually 5 or 10 trials.
Sample
The number of correct trials within a set.
The sample determines PDS.
Split
Extra steps added to the training plan.
When step 1 is too easy and step 2 is too hard, 1.5 splits the difference.
Split Indication
If a 2nd pushed set at step 2 indicates another drop, add a split.
- Push on 1
- Drop on 2
- Push on 1 again
- Drop on 2 again
- Split to 1.5
Criteria Change
how and when
Always by pre-determind PDS.
Always between sets.
Never fudge or fish between trials within a set.
Session
Total time the dog is trained on any behaviors in one go.
Example: train for 30 minutes, including sit, down, LLW, recall. Progres
Economy
The degree to which a particular class of motivator is either earned or given for free.
Water should be a completely “open economy,” available at all times.
Preempt losing the dog’s attention
trainer’s side
- Practice mechanics without a dog to increase speed and efficiency.
- Follow your training plan!
- Train faster! Low ITL/high RoR
- Predetermined PDS points
Rate of Reinforcement
**split card **
RoR
The number of reinforcers delivered to the animal per unit time, usually expressed per minute.
Keep it high!
Wasted time loses engagement—entirely a trainer error!
i.e. 12 treats/minute, or 6 tugs/minute
Define and identify cause
Goldilocks Zone
what and when
Excellent RoR of 8-12/minute
10/min (every 6 seconds) for novice dogs, even if they are very keen. In practice, shoot for 8-12 per minute.
Natural result of a good plan and proper application of PDS.
10/min is ideal. 8-12 keeps us in the neighborhood.
Checking RoR
Check often when working with novice dogs (first ~20 sessions).
- Easiest to use systematic PDS.
- Estimate with treats.
- Count a number of treats available for a series of sets—say 40
- Train for 3 minutes
- Calculate RoR from leftover pieces
- Video yourself and count RoR.
Wasted time loses engagement. Split as needed to maintain momentum.
Increasing RoR
- Smooth mechanical training skills
- Minimal Inter-Trial Latency (ITL)
- PDS rules
- Drop, split, or shop as needed
- Pay setup behaviors (i.e. Sit for down-from-sit)
Using a lower Rate of Reinforcement
when and why it can vary
Only once a dog is a seasoned learner.
- Driveyness to train
- Experience level
- Tolerance for low RoR
- By behavior (stay, long-duration heeling, etc)
Training Flow Chart
Watershed decision based on the dog’s current emotional state. CC or OC?
Yes/left: CC to address upset emotional state. No/right: straight to OC.
What vs. Why
hiccup in training
The first step in addressing a problem during training.
Is the dog not motivated (why) or is the dog just confused about what to do (what)?
Motivation for a reinforcer must ALWAYS come first. Why before what.
Motivation
or lack thereof
No motivation? No training.
If you don’t have control of what the animal wants, you cannot train them and should not try!
Always address the “why” for the animal, before trying to get to the “what” of behavior.
“Why should I?” -The dog, every time.
“Free Lunch” Myth
No properly functioning living organism will do something for nothing.
**split footnote onto card **
Little to no energy wasted on inefficient/unbeneficial behavior.
“Desire to please” is utterly disingenuous. Trainers who cannot address motivation are incompetent.
“Why” Problem
No point in attempting to train an unmotivated animal.
Motivation is the trainer’s problem!
If they won’t take a freebie you don’t have motivation.
Addressing Zero Motivation
long term
If a dog is not motivated by food, toys, petting, or praise, close the economy on one or more.
Why animals behave
according to Dr. Susan Friedman.
“Behavior doesn’t just flow like a fountain. Behavior is a tool animals use to produce consequences.”
Jean’s #1 thing to impart to owners.
Talking about Motivation with Clients
“Dogs do what works.”
- Will repeat many times for all clients.
- Be patient and steady in explaining!
- Motivation is always the keystone of training, even of their past pets.
Sometimes deconstructing past pets’ behavior is helpful, sometimes not.
Leverage
Actively Control the Motivator
What does the dog want right now? Food, play, person, door opening, adventure, smells, etc.
- Instantly start it for R+
- Abruptly end it for P-
Making the dog a “believer”
Training relies on the dog believing that you control their motivators. Builds over time.
Don’t have control? Don’t train!
Avoid situations where the dog may see that the trainer doesn’t always have control, or at least avoid cueing behavior. Gradually, you can train in less controlled circumstances.
Hydraulic Reinforcer
Motivations which wax (deprivation) and wane (satiation). Large impact on power of motivators right now.
Eating, drinking, sex, and play are good examples.
Implication of Motivator Hierarchy
regarding hydraulic motivators
If a hydraulic behavior has been satiated, it may become momentarily annoying or distracting.
Motivations differ throughout the day—use what works right now.
Troubleshooting Motivation: dog takes freebie
RoR is too low, or you have a “what” problem.
Troubleshooting Motivation: dog refuses freebie
Trained for too long, or competing motivators.
Your options:
* End session
* Use establishing and abolishing operations
* Upgrade/rotate/change class of motivator
Short Term Motivation Fixes
Salvage this training session.
Audition other motivators
- Upgrade within a class (i.e. kibble to chicken, or ball to stuffie)
- Rotate within a class if upgrade unknown
- Switch classes—swap food for praise or toys
- Sometimes novelty alone works
Remove clear competition
- More distance or put something away
- Change rooms and close the door
- Visual barrier
Employ the competition
- If you can’t beat ‘em, join ‘em.
- Premack! Make the more powerful motivator your new reinforcer.
Long Term Motivation Fixes
Create motivation for this dog down the road.
Lay the groundwork before the session.
Deprivation
- Establishing operations—establish a motivator
- Example: close the economy on food or skip meal(s) before class
Saturation
- Abolishing operations—make the motivator boring by opening the economy
- Example: play with dogs or cruise the yard for 10 minutes before training session
Pod
antecedent tool
Visual barrier—a valuable tool in group classes
If all else fails, coach handlers on Premack (good for class assistant)
Fade the pod: once dog is performing better, move beyond the barrier without removing it. If it’s too much for the dog, go back behind it.
Economics of Training
Expensive behaviors require generous, valuable rewards.
Recall is really easy with high value reinforcers supplied generously.
Toy Reinforcers
Making toys come to life!
* Throwing/fetch mimics fleeing prey
* Tug is resisting like struggling prey
Do not just hand over a toy.
Drivey dogs
Dogs with never ending motivation for toys (i.e. border collies, Malinois)
Middling drive dogs
Limited interest in toys, like Goldens and Labs.
- Control the economy—no free fetch or tug
- Keep toy-motivated session time short so you finish while they still want more
Non-drivey dogs
Dogs with little prey drive, so toys aren’t reinforcing. Develop their food motivation instead.
Rules, best use, and exclusion
Tug
Be strict! Easy to teach in the course of play.
- Start cue—dog only grabs when invited
- Stop cue—dog lets go on cue
- Accuracy—dog never puts teeth on human
- Breaks—lots of breaks to “test system” and practice obedience
R+ is the re-initiation of tug, P- is the cancellation of the game.
Great motivator for action behaviors like heeling or recall.
Never model tug for small children.
Closed Economy
Never giving a class of motivator for free, thereby increasing its value.
100% of a given motivator is earned in training.
Open Economy
Allowing unimpeded access to a class of motivator, thereby reducing its value.
Criteria
Your contract with the dog. This is exactly what they need to do right now to be paid.
A complete description of current behavior dog must perform to earn a reinforcer.
Ex: Down on hand signal, remain down at 10’ from handler for 5 seconds.
Criteria Quality
Always define parameters!
If you aren’t sure, they can’t possibly know what you want. Follow the training plan!
- Bad criteria: Going into the crate
- Good criteria: Going into the crate from 2 feet away, on a verbal cue followed by a hand prompt (tapping crate)
See Dog Training module slide 34 for more examples.
Criteria Setting
Incremental steps in a training plan.
Always aim for the “appropriate” sweet spot.
* Easy enough to win often (keep RoR high)
* Hard enough to progress quickly through the plan (keep pushing as able)
Best check of criteria—optimal Rate of Reinforcement!
Parameter
Aspect of a behavior comprising specific criteria for reinforcement.
- duration
- distance
- distraction
- degree of prompting
- stimulus control
Increasing Parameters
Generally best to increase difficulty on one paremeter at a time, often decreasing others/another at first.
Example: decrease distance from handler for increased duration.
Inter-Trial Latency (ITL)
what and addressing
Time between trials. Keep it low!
- Move smoothly and quickly.
- Crank out trials as fast as you comfortably can.
- Begin the next trial once a reinforcer or no-reward marker has been delivered so you and the dog are working continuously.
Trainer mechanics are smooth—drill as needed.
Alternative PDS systems
Push: 7+/10
Stick: 5-6/10
Drop: 4-/10
Push: 5/5
Stick: 2-4/5
Drop: 1-/5
Critical: whatever your PDS system, be consistent and stick to it.
Rate-Criteria Relationship
Inverse.
Criteria raised—rate usually goes down
Criteria lowered—rate usually goes up
Discretionary Sticks
- Second guess the indicated Push
- Override your PDS system to Stick
If often, your PDS system and comfort zone are mismatched. Use a different PDS system.
Shopping
simplified freshman definition
Intentional use of discretionary sticks until the dog is already performing at the next criteria step.
“Pavlov is on your shoulder.”
Bob Bailey’s wisdom
Don’t neglect Pavlovian ramifications during OC.
Be careful about aversives because they carry nasty side effects.
Four key elements of good training
Per Bob Bailey
Rate!
Criteria!
Timing!
Mechanics!
Mechanical skills of training
- Set position
- Quiet body
- Prompting & fading
- Timing
- Position feeding
- R+ storage & delivery
Can only be improved through deliberate practice.
Per DT module slide 56. Compare to “technical prerequisites for efficient OC” flashcard
Set position
Standard ready trainer pose. Usually hands behind back, unless using for antecedents or consequences.
No tells before click or reward (besides the click).
Quiet body
No extraneous movement by the trainer.
Lets dog focus on what matters
- Hand prompt/cue
- Hand delivering primary R+
Moving parts should be salient—don’t lose them in unnecessary “noise”
Timing
marking and consequence
Latency between behavior and R+ or P-
“You don’t get what you want, you get what you pay.”
Secondary reinforcer (like a click) are especially important for fleeting behaviors like blinking, eye glance, or nose touches for timing.
Good Timing
- Reinforcement
- Appropriate RoR
- Low ITL
Clicker
- A secondary/conditioned reinforcer
- Allows more precise reward marking of correct behaviors
- Allows more time between behavior and reward for correct behavior
Great for fleeting behaviors like blinking, eye glance, or nose touches.
Types of clickers
- Box clickers—loud with unique sound, may need desensitization, good for distance
- i-Click—faster and quieter than a box clicker
- Tally counter—very quiet clicking sound, counts reinforcers and facilitates rate checks, especially handy when rate dictates the action (example: free-shaped retrieve)
i-Click is Jean’s go-to clicker.
Clicker charging protocol
Classical conditioning—click must 1:1 predict R+
“Click like a statue.”
- Sequencing—click first THEN reach for and deliver treat
- Blocking—reaching, bag crinkle, beginning treat delivery
- Vary time! Random seconds within session or, preferably, randomly throughout day followed by higher value R+
[sleep between sessions?]
Clicker charged dog
- At least two click > treat sessions with sleep after each
- Dog should head-snap for a click
Prompts
Lures, hand signals, or other antecedents which elicit a behavior.
CC between cue and prompt requires a clear separation!
Prompt Jumping
steps for shopping
- Set: Cue > Prompt > Behavior > R+
- B for Prompt is the criteria
- Repeat repeat repeat for CC
- dog will perform the B on Cue (before Prompt)
- Repeat same set until jumping the prompt reliably
- PDS rule—4-5/5 for the beyond-criteria jumping
- Same sequence (Cue > Prompt > Behavior > R+) with pushed criteria
- higher value reward for Behavior on Cue alone
- lower value if prompted
Example of response cost (split R+)—food and praise on cue, praise only on prompt.
Cue sequencing
1) Verbal cue
2) Pause
3) Hand signal/prompt
Often have to let this go with clients/students.
Click sequence
1) Click
2) Pause
3) Reach for primary
4) R+
Target stick training
Value for pet owners
- Easily move the dog in space away from you
- Set criteria
- Timing practice—capture fleeting behaviors
- Tricky coordination
- Super fun for the trainer!
For intermediate to advanced students.
Value in training “stiff” little dogs
- Tough to lure
- Often need splits
- Position feeding must be precise and quick
- Easy to accidentally lure a different position
- i.e. forward and slightly down for Stand so they don’t Sit.
Backdoor Sit and tunnel prompt for Down.
Tunnel Down
Shaping Down by moving their head through a tunnel.
- Start with a high tunnel (leg)—just need to slightly lower head, barely into tunnel
- On 5/5, slightly further into tunnel
- Same distance in, but squish the tunnel a little tighter
- Complete sets changing either the distance into tunnel or “degree of squish”
- Eventually, dog must lie down to meet criteria
- Keep lure low, but fade tunnel/leg by backing it toward the dog
Can typically run through all of the steps in a 5-10 minute session.
Backdoor Sit
For dogs who struggle to Sit from Stand, but readily Down.
- From Down, lure on a diagonal so their butt stays on the ground as they Sit
- Pay in Sit position
- From Stand, use a brief Down to lure Sit
- Down acts as a stepping stone, fading ASAP until Sitting from Stand
Feeding for Position
**break up this card **
Define
Uses
Sit happy position feeding
Feeding to get the dog in a particular spot at the conclusion of the trial in order to set up the next trial or to develop a particular habit.
Consistent delivery of R+ will create a bias so the dog orients or ambulates toward the source after a reward marker.
“Setting up the next shot” in billiards.
For “sit happy” dogs learning Stand, they may Sit while you deliver the treat. If so, prompt another Stand so you can pay with them in the correct position.
“What” Problem
broad definition
A collision between the trainer’s expectation and the actual strength of the behavior.
Reasons for “What” Problems
- Under-trained—trainer presumes “knowledge” based on witnessing correct responses
- Failure to generalize
- Poor inter-cue discrimination
Under-trained
Volume issue—the dog needs more training, aka reinforced repetitions.
Behavior ”should” be better by now.
Failure to generalize
The dog needs more repetitions in different environments.
New place, distraction, handler’s body posture, etc.
Define
Poor inter-cue discrimination
Strong isolated behaviors, but the dog can’t distinguish between cues when mixed together.
How dogs learn
long term
Like dancing or playing an instrument for us—maintained through ongoing practice of repetitions.
There isn’t a moment or flash of insight when they “know” the behavior.
Misleading frames for animal training
Unhelpful ways to conceptualize/phrase animal learning
- “Know/knowledge”
- “Realize”
- “Understand”
- “Got it”
Successful trials don’t mean that
training is done!
People who believe dogs learn in these ways get stuck when the dog gives “wrong” responses after he allegedly understands.
What training achieves
Alters the probability of behavior—strong or weak.
Sand on a scale—reinforced responses are a little more probable later.
Ways dogs guess
- Frequency—cumulative R+ history
- Recency—latest trick syndrome
- Preference— self-reinforcing behavior
- Sequence—repeating a common series of behaviors
Order of events example: sit, down, and roll over when you say “Sit.”
Key to training and testing verbal discrimination
Mix up the order of cues so they aren’t a clue for guessing.
Troubleshooting verbal discrimination
Describe the sticking points and fix for each
-
Signal overshadowing verbal cue
- separate for each behavior in drills
-
Waits for signal
- discretionary sticks for prompt jumping on verbal
-
Poor accuracy/continued guessing
- raise the criteria, imposing a cost for guessing wrong
- Imposing costs says, “Try harder or at least try something on a verbal cue.”
Varying R+ value
when and how
- Higher value reward for behavior on cue alone, or lower value if prompted
- Higher value for pushed criteria in shaping
- Great food vs. good food, food vs. praise only, etc.
Incentivizing pushed criteria over guessing.
Does a higher value for pushed criteria in shaping count?
No-Reward Marker
how and function
Marks the timing for incorrect guesses—gives meaning to the cost
- Respond to incorrect response with “too bad” or “whoopsie”
- Do nothing for a few seconds
- Dog will learn it predicts a crash in RoR
Cute no-reward marker like “whoopsie” is hard to say harshly (become P+)
So it’s a secondary P-, right?
Technical prerequisites for efficient OC
What a good trainer does!
- Good mechanics
- Timing and quiet body
- Sequencing—order of prompts and cues
- Clean prompts
- Smart position feeding
- Solve motivation problems
- Planning and criteria setting
- Incremental training plans
- Clear PDS rules
- Splitting as needed
This one from DT module slide 92. Compare to the “Mechanical skills of training” flashcard.
Optimal session spacing
Meyer and Ladewig’s 2007 study found that once weekly reduced total training time.
More research is needed.
Practical session spacing
Train when you can!
Likely constrained by owner logistics.
List sound mechanics
“Foundation of all animal training”
- Quiet body & set position
- Clean prompts & fading
- Timing & sequencing
- Cues before prompts
- Clicks before reaches
- Position feeding
- Working quickly (good inter-trial latency)
Per DT module slide 94 under “Mechanics.” I added set position (per self-assessment form). Compare to prereqs and mechanical skills flashcards.
Value of training session videos
Opportunity for self-assessment!
- Become technique conscious
- Eliminates uncertainty about your technique
- Catch bad habits early to avoid rehearsal
- Recognize your own good habits
- Better able to break it down and impart it to others—true mastery
Savant-type trainers don’t know what they’re doing right and thus can’t help their clients improve.
