From: nrf (noglikirf@hotmail.com)
Date: Sun May 11 2008 - 01:08:58 ART
----- Original Message -----
From: "Anderson Alves" <mota_anderson@hotmail.com>
To: "'Lucy Favaloro'" <lucyf@cv.net>; "'Colin McNamara'" <Colin@2cups.com>;
"'A.G. Ananth Sarma (GMail)'" <ananth.sarma@gmail.com>
Cc: "'Joseph Brunner'" <joe@affirmedsystems.com>; "'Himawan Nugroho'"
<hnugroho@gmail.com>; "'groupstudy'" <ccielab@groupstudy.com>;
<comserv@groupstudy.com>
Sent: Tuesday, May 06, 2008 3:30 PM
Subject: RE: How to Become a CCIE v2
> Hi Lucy,
>
> From what I understand, I guess a brute force method for CCIE (it's when
> someone takes a lot of lab attempts prior to pass the lab), at least when
> hacking something this is the main idea of Brute force when you want to
> discovery a password you try a lot of different combinations until you get
> it.
> But for me there is no such a thing (a brute force CCIE method may exist)
> but not a brute force CCIE, at the end the person will still be called a
> CCIE just like a Judge is a Judge (no matter how long it took him to get
> his
> position).
Yet the issue to me is not just about simple definitions, but about the
meanings behind those definitions. After all, obviously, with the rules in
place today, anybody who passes the lab is a CCIE even if he needed a
hundred attempts. That's because Cisco has defined it to be so.
The real issue is what it means for the overall integrity of the
certification itself. Employers today simply have no idea if somebody
passed the CCIE because he is genuinely skilled or because he just kept
taking the exam over and over again until he finally got an exam version
that he had seen before or that asked him topics that he knew well (and
didn't ask him anything that he didn't know). That's the problem.
Hence, one of my proposals is to actually make that information public. For
everybody who really and truly believes that all CCIE's are the same
regardless of how many times they took the exam, hey no problem. Cisco can
simply publish on the CCIE Verification Tool how many times every CCIE took
the exam before they finally passed. If that is too 'harsh', then they can
simply use thresholds, i.e. somebody needed "Less than 10 attempts" or "More
than 10 attempts", or something like that. After all, if every CCIE out
there is really the same regardless of how many times they took it, then
there should be no problem in publishing this information, right?
>What kind of quality control are you talking about? Cheating in the
>exam? You don't have to look too hard to find examples of that in
>schools. It would be hard to cheat in a CCIE lab exam.
Actually, it's not that hard at all. There was that one rather infamous
case where one of the proctors was himself caught selling test questions on
the black market. Then of course you hae test-takers who blatantly break
the NDA by immediately posting everything they remember about the exam on
the Internet, or blabbing to all of their friends about what they saw.
>I think NRFs recent point is school has much more quality control, and the
>CCIE does not. It would be a fantastic policy if Cisco only let you take
>2-3
>attempts per year, but what would that accomplish? I'll tell you, lost
>profits.
Hmm, considering that made over $8 billion in net income over the last 12
months, whatever profits they are getting from the CCIE cert process is a
rounding error of a rounding error. And besides, I would argue that
whatever tiny fraction of profits they might lose would be easily made up by
a more robust and more trusted cert process.
But even if profit was the issue, then the answer is simple: Cisco can
simply increase the cost of each exam attempt. Be honest - how many people
here are really going to stop pursuing the CCIE if Cisco doubled the price
of the exam? Or even tripled or quadrupled it? I would argue that not many
would, and certainly the dropoff would not be commensurate to the price
increase (i.e. a doubling of the price would result in less than half the
people stopping. a tripling of the price would mean that less than 2/3 of
the people would stop, etc). To use economists' parlance, the demand is
quite inelastic. Let's face it. A lot of people don't even pay for their
own attempts (they get employers to pay), so they don't care about the cost
anyway. Furthermore, a lot of people who do take the exam now are quite
lavish about paying for high priced bootcamps. If they can pay for those
bootcamps, they can certainly pay for a more expensive test.
>The failure in your logic is that you are comparing a regular class exam
>with a CCIE exam. At least on a regular
>exam you know what chapters to study and your professor/teacher goes over
>the skills you must master in class.
>You may have a quiz on those skills to aide you on which areas you are not
>doing well in so you can concentrate your studies, this is not true for a
>CCIE test.
Actually, no, that's where your logic fails. I am not comparing the CCIE
with just a regular class exam. What I am comparing is the ENTIRE TOTALITY
of the school experience vs. the CCIE. Merely passing a single class exam
clearly does not confer you a degree. You first have to get into the
school - which is usually the most difficult part for any reputable school -
and then you have to pass NUMEROUS class exams before you finally graduate.
>Not only do you have to know the material you
>have to configure it and get it working in a very short period
>of time. I have a bachelors in Mathematics/Electrical Engineering and a MS
>in Telecommunications Managemnet, hands down the the CCIE test was the most
>grueling test I ever took (8 hrs). I studied for approx 8-10 months. During
>that period
>I normally studied about every day, for anywhere up to 4-12 hours a day.(I
>was able to do some studying at work)
Yeah, maybe the most grueling SINGLE test you ever took. But how many TOTAL
tests did you take in college?
Now, where the comparison is more apt is in relation to PhD qualification
exams. Note, these are not the exams you need to actually GET the PhD.
These are just the exams you need to pass to simply qualify to become a PhD
CANDIDATE, after which you still have to complete the research and write the
accompanying dissertation itself, which is by far the hardest part of
getting a PhD. Depending on the school, these are exams that also take at
least a couple of full days, and in some cases, up to five full days to
complete. And practically all of them nowadays includes a day of oral
defense, where you basically have to stand up and answer questions from your
committee about anything they deem is relevant. Your first 2-3 years of
full-time PhD study (hence, you have no job) is specifically devoted to
preparing you to pass those quals.
The major difference, again, is obviously that you can't just keep taking
the quals over and over again until you finally pass. Most schools have a
limit of 2 or maybe 3 shots. Some have a (brutal) limit of one shot, that
is, you fail, and you're immediately expelled from the program. Heck, I
remember one guy who was debating with himself whether he should renew his
apartment lease for another year because he didn't even know if he was going
to pass his quals, which meant that he would have no use for the lease.
{Turns out that he did renew the lease and he did pass the qual, so it was a
happy ending for him.}
But like I said, the CCIE is not like that at all. You really can just keep
taking the exam over and over again until you finally pass.
>On a CCIE exam the amount of potential material to study for is enormous,
>take a look at a blueprint.
Then, as a point of comparison, take a look at the 'blueprint' of things you
had to study to obtain a college degree. Not just one class, but the entire
degree. I think we can agree that the comparison isn't even a close call.
Look, I agree that the CCIE is clearly more difficult than a single (or
perhaps even several) college classes. But that's not what we're talking
about, and that's where your logic fails.
>The other failure in your logic is that you can take the same test over and
>over again. For example,
>the current CCIE security exam has at least 6 versions of the test, perhaps
>more. Its not the same test,
>the change little things as well which force you to know the material
>inside
>and out.
Uh, no, that's not what I said. I never said that you take the SAME test
over and over again in the sense that you get the exact same version every
time. In fact, I explicitly said that this is not the case.
What I said is that as you take the same test over and over again, you will
INEVITABLY get a version that you had seen before. After all, there are
clearly not an infinite number of versions out there. From the simple rules
of probability, if you keep showing up over and over again, you are
eventually going to get a test version that you have already seen.
But even that's not the most serious problem. The most serious problem is -
as you said - because each test version is different, you are inevitably
going to get a version that you know well. That's because every single
person out there has certain technologies in which he was strong and others
in which he was weak. By retaking the test over and over again, you are
inevitably going to get a version that you are more likely to pass simply
because the questions are mostly about things that you know well.
Let me give you a case in point. I'll use an old example so as not to break
NDA. I know a guy who was a veritable master of Token Ring and IBM
networking(I think I can talk about that because those technologies are no
longer on the exam). He was so skilled because his whole job basically had
to do with taking care of a large TR network. And sure enough, when he sat
for the CCIE, he luckily got a version that was very heavy on TR and IBM
questions, and he easily aced those questions, scoring full points. But he
didn't do that well on some of the other questions, and so he just barely
got an overall passing score. He freely admitted afterwards that if he had
gotten a version that wasn't so heavy on TR, he probably wouldn't have
passed.
Now, I don't know this for sure, but I surmise that if he wasn't so lucky -
if he had gotten a test version that wasn't heavy on TR and hence failed the
exam - no problem. He would have simply gone back over and over again until
he finally did get that version. Heck, I would have done that if I was him.
And that's the problem. The current structure simply encourages people to
keep repeating the exam until they eventually get the version that they
want. It's like if the Patriots could just keep playing the Giants in the
Superbowl over and over again, the Pats would eventually be able to catch
the Giants on an off-day.
>The CCIE is also a true test of YOU. Your ability to focus, your ability to
>conquer odds. The quality control is neither given nor expected.
To that I would say that quality control is precisely the problem of the
CCIE. Like I said, I don't see what's so controversial about giving people
a lifetime limit of, say, 10 lab attempts. After all, if you can't pass the
lab after 10 shots, come on, I think Cisco ought to cut you off. More
importantly, it would encourage everybody to be more careful about each
particular lab attempt in order not to waste any of their 10 shots, as
opposed to what happens now, when some people take the test as little more
than a "scouting mission" to see what it's like so that they can make a
"real" attempt later.
Now, of course, we can argue whether the proper number should be "10" or
something else. But the optimal number clearly isn't infinity, which is
what it is now.
But, more importantly, the CCIE is NOT really a true test of you. In
particular, the CCIE process is clearly slanted towards the rich. (In
fairness, schools are too). Specifically, it is the rich who are the most
able to afford to take the exam over and over again. More importantly, it
is the rich who can afford to buy all of the equipment to create a complete
facsimile of the lab in their basemen and to attend all of those bootcamps.
Clearly those bootcamps aren't charities - they only teach people who can
afford to pay. What those bootcamps are basically selling are increased
chances at passing the lab (after all, what are they really selling if not
that?). Poor people are therefore clearly at a disadvantage in becoming
CCIE's because they can't afford any of these things.
Furthermore, the CCIE process is ALSO affected by affirmative action, if
indirectly. I'll give you an example. Cisco itself is a proponent of
affirmative action when it comes to employment; and one of the easiest ways
to become a CCIE is to actually work for Cisco. Why? Because Cisco
subsidizes the lab attempts of its employees. Cisco can obviously provide
extensive practice lab facilities to its employees. {For example, I
remember when the CCIE lab had just implemented the 3900 TR switch, people
were scrambling to even get a hold of one in order to practice, but Cisco
employees clearly had access to it.} Cisco also obviously has the largest
community of CCIE's of any employer in the world, which means that there are
plenty of people around who you can give you tips. The same can be said for
numerous other large employers like EDS, IBM Global Services, etc. The
upshot is that people who work for these companies have significant
advantages in terms of becoming CCIE's compared to people who don't work for
these companies. All of these employers use affirmative action.
This archive was generated by hypermail 2.1.4 : Mon Jun 02 2008 - 06:59:16 ART