[Titus Quinctius Cincinnatus]

You're still not getting it. Do you understand how multiple dating methods applied to the same data sample yield a convergence of the possible initial conditions, which in turn yield their own standard deviation? Geologists who date rocks certainly do. Don't ya think that someone would've noticed this 'little problem' before you and fixed it? Or are all geology journals engaged in a massive conspiracy to destroy the TRVTH of the Magical Young Earth?

Developing a standard deviation by associating independent data sets in this manner is largely meaningless. The simple fact of the matter is that when you date rocks using one method, you arrive at one number. When you date rocks using another method, you will most likely obtain a number which is widely divergent from the first, and on down the line - the isochrons obtained by independent testing using each method are often nowhere near being within the confidence level for each independent data sat. Treating them for convergence is merely a way of getting around this, so that the data sets can be plausibly associated to give a statistically "average" number (used colloquially). It's not a conspiracy by geologists to hide anything, it's just a way to make the numbers smooth out. Yes, okay, you get a number, and statistically it works out, but it's not "real", which is the point. It's a statistical construct which only has relevancy because it makes everyone's lives easier.

None of this addresses my original argument, however, which would basically be that you can collect all the data you want, using whatever methodology you like, but if your starting assumptions about how the rock "acts" and what it's condition was when it first formed are faulty, then any subsequent data collection is ALSO faulty, regardless of how you statistically treat it. If you don't really know that the rock had no argon (for K-Ar dating, e.g. - and there is research out there that seems to indicate that we CANNOT make that assumption) to begin with, then you can't confidently say that you know what's its age is NOW based upon the decay rate of ⁴⁰K. This precedes statistical treatment, which is what you're missing.

[ok_now]

Yes, okay, you get a number, and statistically it works out, but it's not "real", which is the point. It's a statistical construct which only has relevancy because it makes everyone's lives easier.

Um, no, your interpretation of the data is faulty. A divergent value for the age of the isochrons is obviously indicative of initial conditions that can't be used. If the isochrons do converge, well, then you obviously have a datable sample. Tell me - can you find any instances where a pyroclastic flow underneath a sedimentary stratum (which generally can't be dated radiometrically) containing benchmark fossils from any era prior to the Cenozoic that returns a convergence in ages significantly more recent than the Cenozoic (barring regions where an obvious upheaval has occurred)? Now that would be impressive news.

None of this addresses my original argument, however, which would basically be that you can collect all the data you want, using whatever methodology you like, but if your starting assumptions about how the rock "acts" and what it's condition was when it first formed are faulty, then any subsequent data collection is ALSO faulty, regardless of how you statistically treat it.

Actually, that's been addressed, you just didn't catch it. There is no way you could get a convergence of diverse isochronic data under any feasible circumstance unless the data meets the initial assumed circumstances. That's where your misunderstanding of the process (no doubt fueled by creationist propaganda) comes in.

Also, it's not like radiometric dating is the only process that confirms the earth is older than 10,000 years old, for crying out loud. (Funny that processes used for dating the age of the sun, which are totally different than those used to date rocks, return the same age for the solar system, Hmmm......)