German Shepherd Dog Forums (http://www.germanshepherds.com/forum/)
- Training Theory & Methods (http://www.germanshepherds.com/forum/training-theory-methods/)
- - There are always consequences... (http://www.germanshepherds.com/forum/training-theory-methods/459994-there-always-consequences.html)
There are always consequences...
To the way you train..
Just going to share one of the speedbumps I have encountered with the current dog.
For those of you that are training nuts and get out there regularly this will come as no surprise to you..still its always amusing when you realize you shot yourself in the foot..lol.
I train my dog both for sport and real life. Sometimes that create conflicts down the road.
I trained my dog to heel "pet heeling" mainly through positive punishment (prong and E Collar) in conjunction with + reenforcement (ball). Dog must stay on my left leg, pivot to remain straight, no forging or lagging. Beign out of position = punishment. I did not install an automatic sit as I did not find it important as long as the dog remained on my leg.
I trained Ipo style heeling first with food then with ball minimal to no compulsion. Dog has decent positioning and heels well when motivated.
However, I began to experience an issue with extended focus and expression (dog started looking across my body instead of at my face / shoulder). This obviously looks sloppier and inevitably leads to the dog completely looking away from you at some point.
Mode of address:
-Increase the number of reward instances as well as vary delivery and ask for less duration between rewards
Result: This did offer a slight increase in motivation but after the first two repititions the issue began to crop up again.
-Create better upward focus through maintaining the reward delivery only from the armpit. (At the time I had begun to deliver the ball from different areas)
Result: Again no improvements were noted beyond the first couple of repitions.
-So now we resort to positive punishment. Punish the dog for anything less then full focus on the handlers face. (Here is were I find the bullet hole in my foot) I go to my two favorite tools prong/ecollar. Dog starts slacking in the focus I correct.
Result: Dog began hugging my left leg even tighter, no improvment in the dog actually maintaining focus on my face. If anything we began to lose even more expression..
Lightbulb: I have corrected the dog in the past for coming off my left leg, forging, lagging etc. In these instances getting back into the correct position removed the pressure and gained reward.
Dog is being confused by the application of punishment as he deems herself to be in the correct learned position which she is. (What I am correcting for is the dog not giving me her full focus) she has learned that in the instance of a heel corrections come for being out of position and have no connection to focus.
Now I know that the way I trained the "pet heeling" is causing confusion when it comes to teaching the IPO heel.
So we go back to positive methods and I think about it for a while. I try to get more ball drive because I percieve this to be a motivation issue. More drive = more desire to perform the behavior correctly = better duration and better expression.
While Im dinking around with this I learn that the dog is more motivated to access the ball when it is lying on the ground or when I actually throw it for her.
Clearly positive only methods arent getting me quite were I need to be time to go back to punishment to make my communication clearer and create more duration, better focus and good expression.
Solution: Place the ball on the ground close to the dog. Call the dog to "FOOSE". Dog comes into position, looks at me = reward (she gets to access the ball). Do a couple times add a bit of duration, rinse and repeat.
Now the dog starts looking away from me because the ball which is lying on the ground is a big motivator.
Correction = the dog looks back and is immidiately rewarded.
Now we add some movement dog wants badly to look at the ball but everytime she does = correction. Now she associates positive punishment with not only her position but looking away as well.
Dog has a lot of motivation for the ball just lying on the ground, we can also increase motivation by tossing the ball and calling the dog to "Foose" while the ball is creating heavy distraction through its movement.
Now since this is similar to how I trained the "pet heel" and I did not install an auto sit we have begun losing the auto sit in the IPO heel. So having to fix that too.
Overall the clear delineation between the punishment and reward was achieved by simply placing the ball on the ground and creating two options.
Look at the handler and you get the ball.
Look at the ball and you get a correction.
We are making up for some lost time now with our IPO heeling and exression has not suffered as the motivation for the ball is very high. I can also now carry the ball and punish for a lookaway and recieve better focus so the ball does not have to be on the ground all the time anymore.
We have someother things we are working on but this specific issue has been annoying me for a while.
So many things. Have seen dogs basically become afraid of the reward with what you are doing. She doesn't understand true attention. Ball under armpit-- look at ball. Ball in the ground-- don't look at it or get zapped. Dog is more motivated if you aren't attached to the ball. Lotta confusion and conflict.
I had a feeling some might try to make things out that way..:).
Dog has no fear of any of the reward objects I use. If anything the drive and expression has now improved. Clarity was achieved through conflict in this case.
A lot of my training with this dog has been influneced with what the dog is genetically and finding what motivates her in various situations.
I know how to apply punishment and reward. ;)
Since talk is cheap here is my dog being afraid of the reward..while being taught "pet" heeling.
When you give the dog the proper foundation and teach the dog the desired behavior, then also condition the dog to work through pressure you avoid supersticious associations.
You might have also considered putting the eye contact on cue. People say watch or look for eye contact and instead of no do an aaaah aaah or something like that as soon as the dog looks away and then use negative reinforcement pops or stims till they get eye contact again then good and keep going or yes and reward. Then the dog knows exactly what it was doing wrong and things are more clear for when you switch to positive punishment.
I think there is great wisdom in admitting and learning from our mistakes. You saw the issue and figured out a different approach to start again, nice :)
The video shows several things,
1. Communication from you to the dog is not clear and pretty grey.
2. Practicing too many actions (going forward, turns, backward motion).
3. Drive building activity is not part of the practice session.
1. Reward from stationary heel position (but only for correct position, body/shoulder/head).
2. When rewarding, stim the dog using momentary nick.
3. Start 1 step motion, only reward in perfect position (body and head), nick at reward. Repeat for several weeks. Now take 2 steps, criteria remains same.
After several months the dog will be able to heel longer but with good position (body and head). Always nick when releasing/rewarding with ball.
Start taking 1 step back, only reward perfect position, nick at reward. Take 2-3 steps back, only reward perfect position. Now take 3-5 steps forward and immediately stop + take 1 step back, nick/reward. The taking 2 steps back and rewarding becomes the drive building activity that you will be using moving forward (reward will switch to less during forward motion and more when moving back). So when you feel dog is loosing oomph, take 2 steps back/reward/nick. Dog perks up big time.
Dog will understand what is required in terms of position, will also very clearly associate nick with perfect position and reward. Months down the road, you can THEN use nick to correct and the dog will NOW know what to do (get in perfect position from nick) as you have ingrained that it will result in reward, drive will go up instead of down as this method of rewarding with a nick increases drive when using electric.
The vid was just to show the dog is not afraid of the ball. Im not teaching ipo heeling in that vid. Therefor precision and expression are not what Im focusing on. Just the dog staying on my leg.
Im training the IPO heel somewhat differently then in the vid. That being said I find your method of adding stim before the reward interesting. I have never used it before however I can see how it can play a role in loading the dog instead of just being pressure. Im not quite clear on why you add a backup before the reward. Obviously overtime the dog would perk up on the backup if thats how you always reward.
So overtime I am assuming you back up to load the dog then move forward again instead of rewarding?
I may start adding the stim before the reward and see how it goes, however in this case the stim has already been used in the heeling context as a correction. So this may cause some confusion if introduced at this stage.
Dont think I need to go back to stationary positioning as that is solid. Even moving position was good it was just the head set that was bothering me.
Ill post a vid of how Im training the ipo heel and you can judge that. Obviously the pet heeling I did early in the dogs life is causing some confusion now which Im ok with.
For the next dog Ill probably train the pet heel on the right side.
The moving back with reward or moving through legs on every step with reward or moving between legs with reward are all exercises that we do to motivate the dog, it keeps things fun and exciting, builds drive, improves position and prevents dog not wanting to be with handler as there is no fun in it for the dog/team.
These exercises also give you a tool that you can rely on at any time to bring on more horse power in OB. The other way is to use electric for everything, dog goes lower and lower in drive, does not want to be with handler as there is no fun. Dog does not know what he did right or wrong as rewards are inconsistent, punishment is everywhere. Many such teams say their OB is flat and dog is not "sporty" but "real", well duh!
Just 2 different ways of training.
Its an over time thing. Soon the ball loses its appeal when its become too associated with the correction. She has great object drive. Instead of correcting for her focus on the ball which was taught inadvertently. She still doesn't understand position and I doubt that she understands true attention. I wouldn't reinforce her preference for the ball away from you but that's a preference. I want them to seek drive satisfaction through me with a dog like that which has that level of object drive. In your "pet heeling," you are still allowing a sloppy version of behaviors you will want more refined for competition. Confusion for the dog. Its not an issue of positive only methods here. Its a problem of not clearly setting the criteria and moving too far too fast. You said your two favorite tools are a prong and e-collar. Where was it positive only in her training?
|All times are GMT -4. The time now is 05:05 PM.|
Powered by vBulletin® Copyright ©2000 - 2016, Jelsoft Enterprises Ltd.
User Alert System provided by Advanced User Tagging (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.