I've not been an overclocker for very long. As I recall, I attempted my first overclock in very late 2006, or maybe even just at the start of 2007. In fact, I didn't even build my first custom PC until May of 2006. Perhaps some may then revoke whatever credentials I can lay claim to.
"What do you know?"
"You haven't been at this game for long enough."
"You haven't earned your dues, son."
I suppose these are reasonable challenges to my credibility as an overclocker. But these are things I freely admit to anyway; why bother disclosing these facts as I have? To not acknowledge that I do not know everything about the subject of overclocking is to indulge in that most dangerous of indulgences: Hubris.
However, I must say that the sum total of what I do know about PC construction, system tweaking, and overclocking were all earned through experience. I've felt many highs (I remember vividly being very gentle with my A8N32-SLI Deluxe when I took it out of its box for the very first time, and then inserting my San Diego 3700+ in the ZIF) and seen a few lows (witnessing firsthand one of the worst things of what an unstable overclock -- traced to bad RAM -- can do, wholesale OS corruption). I've had to RMA a couple of items (RAM and motherboard), learned that some apparently great things (SLI) are more gimmick than perceptible performance advancement, and felt the thrill of a full 50% overclock on an Opteron 165 (1.8GHz to 2.7GHz, a gain of 900MHz).
There are still some mountains I've yet to climb. I still haven't overclocked my graphics cards, I haven't done any water-cooling yet, but these are challenges for the future. The only reason I've not tried these things out yet (well, besides than committing to a budget for these, since replacing an OC-damaged video card costs a lot of scratch, as does a properly-designed and executed water cooling setup) is that I've not amassed enough knowledge on these subjects. No knowledge, no confidence.
What baffles me sometimes, though, is something I've been seeing far too often in many of the overclocking community forums I frequent, and that is a very reckless brand of advice-giving. What do I mean by this? Well, I'm sure that you've all read posts recommending this setting or that, or someone who says "your CPU should be able to do this speed," or the like. To be completely honest, this is a very dangerous way of working because there is no apparent analysis going on.
Overclocking is not mathematics, though it deals with numbers; it is not science, though it is of course subject to all physical and chemical laws. It is called a "dark art" for good reason: Though there are definite techniques to use, and a method to the madness, overclocking is more like jazz music than, say, disco. Jazz is all improvisation, feel, and intuition; disco is a robot, a mechanized simulation of music. Live music is all about the moment, as opposed to a studio recording, which is a frozen moment of time but one that can be repeated forever.
Overclocking is a highly contextual thing. This means that my setup is unique and can never be directly comparable to yours or his or hers. Even if we have the exact same types of components (for argument's sake, we have the same model of motherboard, the same model of CPU from the same stepping batch, the same type of memory, the same HDDs, the same PSU, the same all the other parts of the equation) in our systems, there's simply no way to know how well these systems will overclock until you've actually done so. I can almost guarantee you, though, that one system will run faster than the other one.
One system will run hotter than the other; one system will be more stable than the other; one system will be cleaner than the other.
If there's one essential lesson I've learned about overclocking, it is this: The only guarantee about the activity is that there are absolutely no guarantees. You either know something, or you don't, and the only way you come to know something is if you've actually done it, tested it, and verified it scientifically.
I understand the spirit behind OCN and other similar OCing forums is to be helpful to as many people as possible. But don't forget one thing: What works for you may not work for whomever you are advising. It's far better to explain why something may not be working, instead of just saying "do this" or "set X to Y MHz." Analysis is the more useful skill; anybody can go into a BIOS and change settings.
Learn as much as you can, even if it is always costly (in terms of time, effort, and, sometimes, money). Most of all, please learn to analyse problems. Understanding and wisdom, after all, are the sweetest fruits of experience.
As always, I invite comments and discussion.