What toy should I use to train my dog?

Rocko's session breaks play into elements the handler can compare: chasing, grabbing, contesting possession and winning. The trainer explains why he thinks contesting possession interests Rocko more than repeated misses during chase. That is a working interpretation of this session, not a standardized preference test or a finding about every dog. Watch the sequence and the dog's response alongside the explanation before applying the same interpretation elsewhere.

The shared owner lesson discusses how opportunities to win and the structure of a game can affect engagement. Kodak's separate session supplies a practical invitation to bring and try different play options: the trainer says trying tug, a ball or another option can change the approach. These passages support observing the individual response instead of assuming in advance that a dog will or will not engage. They do not, by themselves, establish the exact minute a session plan changed or prove one explanation for that change.

Deuce is where the finding is reached three separate ways and still surprises the trainer. It comes by telephone first - the owner is asked what he plays with and reports that the ball is not it but tug is. It comes by testing in the field. And in the end it comes simply by watching what he does when the work stops: he walks up and asks to be patted. The trainer says out loud that this is not what he expected of the breed. His file also holds the honest footnote to the whole idea: at the handover his trainer says he could never get the recall to produce the speed the ball produces, which is a limit of the training stated in front of the family rather than hidden.

Sage is the dog who will not take a toy, and her file is where the principle is generalised past toys altogether. She does not want treats and does not want much in the way of games; what she wants is the couch, personal play, and free time, and she chases a rope only outdoors. The line her senior trainer uses is not a decision but a reading - she is telling you what her reward is, so use the couch every time you break her out of the command. The owner lesson turns that into the general rule, and it is the sharpest statement of it anywhere: people have a one-track mind about rewards, usually treats or praise, when there are hundreds of positive reinforcers available; so instead of arguing with the dog you observe the behaviour and pick alternatives, because what motivates a dog is individual.

Dutchess is the case for touch as a first-class reward rather than a consolation prize for a dog who will not play. Tug was tried and reported honestly as not much of a motivator; what actually moves her is a rub on the face immediately after a command, and lazy time on a bench. Withhold it and the work degrades - if she does not get enough love she stops doing anything. And her file adds a criterion nobody else in the corpus names: cost to the owner. Rubbing a dog is free, needs no equipment, is always to hand, and does not wear out, which matters because the family has to keep paying it after the trainer has gone. The theory behind it is taught to her owners at lesson one, from the punishing side: petting a dog that hates being petted is a punishment, and every dog chooses for itself what counts as reinforcement.

Bullet is the equipment lesson. Day three is a play assessment in its entirety - a rope toy, a ball on a rope, a helper showing it and throwing it, the same thing again on a long line - and it ends in a recorded negative: he does not care about any of the toys yet. Three days later a solid toy goes in and he takes it, and the reading is not that the drive appeared but that the object had been wrong all along, because his bite is stronger than the tug was. That single sentence is the practical form of this whole concept, and it is the one that turns 'my dog isn't into toys' from a verdict into a question about what is in your hand. His file also carries the qualification that stops it being a universal fix: a fortnight later the same dog is diagnosed as never having played at all, which is a different problem that better equipment does not solve.

Louie's first session is the textbook version, including the negative result. The handler offers a ball, tries to loosen him up, tries a rope toy, puts him on a line so he can move, and tests whether the line itself is what is blocking the running - and then stops, records that none of it worked, and says the next step is ordinary walks before he can expect engagement at this level. Later files find the answer, and it is not a toy: the paddling pool, and jumping up to be petted. Two motivators, rotated, with the pool doubling as the return to baseline between repetitions. The line that sums it up is as blunt as the concept gets - anything as a reward, whatever it takes.

Lola is the case where the reward has not been found yet and the trainer says so rather than pretending. Mid-programme he watches a perfect down come out of sustained leash pressure and draws the honest conclusion: she knows the position, so the missing thing is not knowledge, it is a reason. He wants something she goes crazy for and has not got one, and he accepts that it may not be findable until she opens up further. What her family does find is a part of a game rather than a game: she loves being chased. The lead trainer's answer is that there are different parts and you can indulge the one she likes - which is the prey-cycle idea arriving in a living room, from the other direction.

Finn is where the assessment fails, and the file is more useful for it. Ball, bone, water, chase, stick and personal play are each offered across several sessions. Personal play makes him freeze rather than switch on - a grab feels to him like something being done to him. What does start him is not under the handler's control, so it is rejected as a training reward on the practical ground that you cannot produce it on demand. At the store trip the trainer says outright that he has not figured out what Finn likes and will have to keep experimenting, and names the consequence: with nothing to fall back on, social pressure has no alternative when it is not working. The eventual answer is movement itself - running with the dog, and tagging the handling he cannot yet take onto the good feeling that running produces.

King is the easy case, and it is included because easy cases are where the rule gets stated cleanly. He takes anything - ball, tug, whatever is going - and the ball is named flatly as his strongest reinforcer, above food, above petting, above praise. That is what lets a senior trainer use him to teach the general shape of the idea to a junior, and the general shape includes the caveat: mixing play into the work makes most dogs enjoy it more and learn faster, and then, in the same breath, some dogs simply do not care for playing. His file also contains the purest version of the assessment, run by a junior handler with no senior present: given two minutes of nothing, King goes to look at the other dogs, and the handler decides that checking out the other dogs is the reward today.

Zeus is the fastest assessment in the corpus and the one that determines a whole programme in a sentence. A heavy tug goes in, he commits to it, and the trainer says it out loud - this is what he loves. The useful part is the second observation, made on a different day: it is not that he likes tug, it is that he keeps possession of it, so possession games become the vehicle for teaching him everything. That reading also generates the work: a dog who wants to keep things has to be taught to give one up without treating the request as a challenge, and the tug is chosen for that lesson precisely because it is low-stakes.

Roxy's household is where the assessment is run on two dogs at once and comes back with two different answers, and where the facility's own process is visible: a handler says the board told him to try fetch and try tugging, and reports both as negative. For the housemate the conclusion is that he likes to walk around; a later session offers the weather rather than the temperament as the reason he will not play. For Roxy the answer is the pool on a hot day, and, on the walk, freedom - and the lead trainer makes a point the rest of the corpus does not: the ranking flips indoors, where freedom is not on offer and the tug or the fetch becomes the best thing available.

Bo supplies two things. The first is a negative result held open for months rather than resolved: the ball is graded honestly as only half wanted and named as a project, and five months later it is still a project. On a walk the environment beats everything in the trainer's hands, and he uses the environment instead - which is where the corpus's best line about reward priorities comes from, said casually to a stranger on a trail. The second is the only piece of reward engineering in the corpus. Bo starts stopping short on the recall, and instead of correcting the stopping short the trainer changes the payment to one that cannot be collected at a distance: jumping up onto him. The last few feet close because the reward is on the other side of them.

Dolce's assessment is run for a purpose the other dogs' are not: her bite and chase play were suppressed as a puppy and the protection work waits on reopening them. So the offering is systematic and the findings are specific. Indoors a tug is asked for first and then a duck, and the duck is what gets her going. On the flirt pole the chase is worked off the back of the possession game she already wants. And on a long walk in the heat her handler names the single most useful reward available that day and it is a drink of water - offered as a general rule, that you use whatever your most powerful motivator is, and that the recall in particular needs a powerful one.

Luna's file contributes the version of this for a household rather than a facility, and it contains the corpus's clearest argument against defaulting to food. Her reward is sniffing and exploring, chosen because that is visibly what she wants. Treats are deliberately not relied on, and the reason given is about what the family is left with: get into the groove of paying with food and the dog is only trained while you have food. Where the family does use treats, the instruction is to randomise them rather than give them on a schedule, because she is clever enough to work a pattern out.

What toy should I buy for training?

None yet. Run the assessment first, because the answer is decided by the dog and the corpus is full of owners and trainers who guessed wrong. Offer, in one session, a ball, a tug, a flirt pole or a rope, a rub on the face, and simply letting him off to go and sniff, and watch which he goes back to. In this corpus that test came back 'a paddling pool' for one dog, 'being patted' for another, 'a drink of water' for a third, and 'checking out the other dogs' for a fourth. Then buy for the answer. If it is a toy and the toy fails, try a tougher one before deciding he has no drive - one dog produced nothing for five days on a soft tug and a hard grip in a minute on a solid one.

My dog doesn't like toys. What do I do?

Two things, in order. First check whether it is really the dog: the object may be wrong, the day may be too hot, the place may be too distracting, or the game itself may be badly run. The trainers give four ways a person makes a game unfun - you never let him win, you win every time without effort, it is exactly the same every time, or you are not enjoying it either and he can tell. Second, stop looking for a toy. The reward does not have to be an object at all, and for several dogs here it was not: the couch, a rub on the face, a pool, a drink, being let off to sniff. One handler's summary is as broad as it gets - anything as a reward, whatever it takes.

He chases the ball but won't bring it back. Is he doing it wrong?

No, and this is where the useful idea in this concept lives. A game is not one action; it is a sequence borrowed from hunting, and your dog is probably buying one part of it rather than the whole thing. The parts named on camera are the chase, the grab, the fight over it, the win, the celebration afterwards, and being chased while he holds it. Work out which part he is there for and build the reward out of that one. If he loves keep-away, play keep-away. A trainer teaching a family with exactly this dog tells them people get caught up in he-has-to-bring-it-back, and that there are all kinds of ways to play the game.

How do I tell which part of the game he actually likes?

There is a test and it takes about two minutes. Play the game and start making him miss - keep the toy just out of reach, repeatedly. If he is there for the chase, missing will wind him up further and the arousal will climb. If it does not climb, chase is not the draw and you should look at the possession fight or the win instead. That is exactly how the reading was made on camera about one dog: he keeps making him miss, nothing much happens, so the conclusion is the fight is the part he wants. The same trainer adds that once the dog has won, watch what he does next - this one just dropped it and lost interest, which tells you the celebration is not his either.

Can I just use treats?

You can, and the trainers here largely do not, for a reason that is about you rather than about the dog: pay with food and you tend to end up with a dog who is trained while you have food and untrained when you do not. Where treats are used with a family, the instruction is to randomise them rather than give one every other time, because the dog will work the pattern out. And there is a practical point behind the preference for play, touch and freedom: those are always in your pocket. One trainer chooses a rub on the face over the tug partly because the family can afford to keep paying it forever, which a toy the dog chews through is not.

My dog is nervous. Should I even try play with him?

Yes, and the corpus contains the case that makes the argument. A fearful dog was brought out with toys in the bag by a trainer who half-expected play to frighten him; four minutes into the session the plan was abandoned because the ball worked, and the dog spent the rest of the outing staring it down. The rule that came out of it is short: carry the stuff, actually try it, and do not let the label decide for you, because assuming a nervous dog will not play is what keeps it true. The honest other side is in Finn's file: with him, everything was tried and nothing landed for months, and the handler said so on camera rather than inventing a result.

Does his breed tell me what he'll want?

Less than you would think, and the corpus keeps catching itself out on this. A big Rottweiler-type who was expected to want the toy turned out to want to be patted, and his trainer says so with audible surprise. A Rottweiler bitch whom the breed would predict to be tough on the e-collar was unusually sensitive to it. Breed does show up as a rough timetable - one handler says he can get a retriever type heeling in two sessions and a working shepherd will not be a two-session dog - and one exchange treats a particular bite item as suiting shepherds and malinois rather than bully breeds. So breed is a hint about the shape of the work. It is not an answer about the reward.

Deciding what the reward is instead of finding out. The trainers' name for it is a one-track mind: treats, or praise, and nothing else considered.

Assuming the breed answers the question. It does not - a Rottweiler-type who was expected to want the toy wants to be patted instead.

Insisting that a game has to be fetch - that he has to bring it back - when the part he actually likes may be the chase, the fight or the keep-away.

Running a game the dog cannot win, or one he wins every time without effort, and then concluding he does not like playing.

Taking a fearful or nervous dog's label as the answer and never offering the toys at all, which guarantees the finding.

Paying with food by default, which leaves you with a dog who is trained only while you have food in your hand.

Ruining the game you just found, by taking the toy away every fifteen seconds or by asking for too much obedience inside it.

You have bought several toys and he has ignored all of them, so you have decided he is not a toy dog.

He will chase the ball but will not bring it back, and you have been told that means he does not play fetch properly.

He works for food in the kitchen and ignores you completely in the park, and you think the training has failed.

He comes when called but stops a few feet short of you, every time.

He is nervous or fearful and you have never offered him a toy because you assumed it would be too much for him.

He plays happily on some days and refuses on others, and you cannot see a pattern - the days he refuses may simply be hot.

He does everything you ask in the house and nothing outside, and you have concluded he is being stubborn rather than that the payment has changed value.