David Hershey schreef
So, how accurate are ring counts for the bristlecone pines? Would you
estimate ring counts are within say 1% or 5% of the actual age? Or are
they less accurate?

The Guinness Book of World Records reports that the 'Prometheus'
bristlecone pine had 4,867 rings when it was cut down. So should
Guinness report a range for the age, such as 4,876 plus or minus 5%?

+ + +
I don't know how it is for bristlecones, but I assume such high profile
cases would be very accurate. Dendrochronology does not really count rings
but studies patterns of growth rings. In a good year a tree will make a good
ring, in a bad year a narrow one. This means certain time periods result in
recognisable patterns: five good years, one bad year, then one good year,
two bad years, one good year, etc. Such patterns can be matched from one
tree to the other, finally resulting in great accuracy. This also means that
trees three hundred years dead can be matched against living trees on the
one hand and trees six hundred years dead on the other hand. The final
accuracy depends on sampling size, tree species (oak being the most easy):
also results are for a limited area only, or rather a certain pattern will
be most reliable in the area where the sampling took place, and will fade
with distance.

Here in the Netherlands, there is a reliable series for oak going back to
the last Ice Age, which means that an oak beam from a building can sometimes
be dated accurately: "this oak was felled in 1248 AD". If a piece of oak was
from the Baltics, this accuracy changes. For other species than oak work is
ongoing, with varying results.
PvR