TRAINING PLATEAUS and OTHER
GENERAL TRAINING IDEAS
by Gail McCarthy
The following summary was written in
response to an inquiry on the SARDobes List where a fellow handler's dog was
ignoring her and not completing the trained indication. The handler was
asking for feedback about how to build focus in her dog......................
Subject: Training Questions
Reply from: mccarthy at macrt org
Just to clarify, you have a "stay-put
and bark live find" dog correct? When you say that your dog "blew you
off," what did he not do? Did he not search in response to the Go Find
command? Did he find but not alert? Did he just race off and sight-see?
Focus is clearly what we all are looking
for as a training objective -- without this we have dogs that just bumble into
scent and may or may not alert....but a search dog should do more than that --
they should SEEK OUT scent, at least in 20 minute intervals, and reliably alert.
Without knowing what he was doing (ie bark
barrel, versus blind inaccessible rubble area versus runway into know hole,
I can't give hard training suggestions but I do have some general things that
I'm sure you, as a sophisticated handler know already:
1) Dogs do not wake up one morning and
decide to not work properly.
2) Dogs are not abstract thinkers and they
are not calculating.
3) They are, however, self-centered and
will willingly repeat behaviors that FEEL GOOD to it and which satisfy some
4) So, if your dog was once performing
well, and now is not, something has changed to make doing the target behavior
not worth his while.
5) Some classic things that make a dog stop
progressing or performing as well as before:
- *Something physical -- injury/sick --
*Dog needs opportunity to run around and run off some steam before being
asked to focus in and work (many horses are like this and Banner needed to
about 15 minutes to get his "yaya's out" before I could start him
in a search area).
*Handler has gotten bored and dog picks up on lack of interest.
*Where once the handler rewarded a given behavior with 3 Big Mac's, handler
now rewards same behavior with unappetizing kibble. We all fall into this
trap but what happens is this "poor reinforcement" makes the dog
less and less interested in doing the target behavior with the risk that
other behaviors (like "racing around and sightseeing") can creep
in and become more self-rewarding (ie, that self-centered cost/benefit analysis again).
*Even if handler is rewarding with luscious Big Macs, Big Macs can start to
become hohum too if that is the ONLY thing the handler is rewarding with.
Remember VSVR (variable systems of VARYING REWARDS) -- if the dog always
gets steak then steaks starts to lose value.
*Again, not being abstract thinkers, dogs do NOT get turned on by the
thought of working problems that are constantly getting harder. I have seen
MORE DOGS QUIT when worked by handlers that say "their dogs need to be
challenged to make them work harder." Nonsense -- no matter how
wonderful the reward, if the dog starts to realize that this search stuff
just gets harder and harder and if the dog starts to connect his handler
getting irritated when he can't make the find within seconds, a dog will
quit. That is why we all need to ping pong problems in difficulty -- and
keep a ratio of about 5-6 easy motivational reps for every problem that
drains the battery.
*Another really important reason that dogs "stop learning" or
"hit plateaus" or "seem to go backwards" is a training
mistake called "LUMPING." Where initially with beginning dogs, we
"split" the training into small component parts and heavily
reinforce those parts, we don't do that when we, as handlers, decide the dog
is "ready" to move on. At this point, even if we do it as part of
a plan, we tend to add in too many variables at once.....ie, LUMPING.
According to behaviorialists, lumping is a major cause for so-called
"learning plateaus," behavioral regression (by asking for too much
too soon), and other such to-be-avoided conditions. To avoid this, and to
get the proper behavior to happen, split the behavior down into smaller
parts with a HEAVY reinforcement schedule so that the dog can once again
feel successful and satisfied. Well, I could go on -- but suffice to say
that, as with Phoenix's refind, we all need to back up with beginning dogs
when things fall apart and return back to whatever point the dog WAS working
successfully. And, JSYK, as part of that directed operant conditioning thing
I was talking about before, I define not only what the target behavior is
but I also clearly communicate what behavior I don't want. And then I
heavily reinforce when the dog performs the target behavior.