Originally Posted by
FBinNY
I think it's clearer if we separate the issues of method from the results.
We all agree that anything between the minimum and maximum (or a fudge version of the maximum) is OK.
As for method, the big/big method is safest because it ensures a chain above minimum. However, nothing says this is the best length. As posted, there are advantages to a longer chain, and the small/small method works fine AS LONG AS you're working within RD capacity.
So for newbies, they should start with big/big and note the maximum number of links they can cut, then use the small/small method to decide how many of those they'll actually cut.
As a rule, those running at or above RD capacity will use the minimum length. Those with capacity to spare, may prefer going longer.
I'm on board, but I think use of the word 'maximum' to define the length of a chain is dangerous. It's a method that works fine 99.9%, but when it doesn't, he results are disastrous (and expensive), and
its most decidedly NOT a 'maximum' length. To me maximum implies that all the other bases are covered, not that there's still a risk of ripping the hanger off your frame.
I've noticed a lot of bicycle mechanics subforum users are recent purchasers of used bikes with unknown modifications to gearing, I fear the subtlety of your bolded caveat might be over the heads of members that actually need advice.
If it's a methodology we're chasing here, something on the order of a checklist/flowchart might be better. I feel the min/max terminology undermines the huge disparity in consequence. There needs to be a paramount priority on big big working as a gear choice, no matter what other criteria exist.
I.E.
do you have the old chain?
Y: match that length with new chain
N: see if big-big works
...
Try small small
...
If small small has flopping issues, take out as many links as you can, while keeping big big working.