Wulf Ratbane
Adventurer
Why are the the dragon CRs multiplied for the purposes of your monster statistics?
They're multiplied in order to keep their data from "polluting" the averages where they were originally located.
It's mostly common knowledge that dragons are intentionally underestimated, so some adjustment is necessary, but I would have expected an addition. Your calculations imply that higher-CR dragons are more underestimated than lower-CR ones; that might hold to an extent, but I don't think it reaches the levels suggested by your calculations at the top end.
It sounds like you understand the methodology but disagree with the specific application.
First, I generally accept the CR calculations done by Craig Cochrane (Upper_Krust around here). You are welcome to disagree, but I do state my biases up front.
Second, the adjusted CR isn't used for any purpose other than grouping the data. I do not recommend changing the CRs on dragons and using it (for example) to price out encounters. If the MM says the dragon is CR12, then use CR12.
But when you're looking at the data in Chapter 1, realize that every monster listed from CR26 and above is (by definition) a data set composed entirely of dragons.
For example, a CR 24 (by the MM) dragon will be a tough fight for a 20th-level party, probably tougher than the CR alone would suggest, but I don't think the functional CR is 24 * 4/3 = CR 32 (party level + 12, practically certain to result in an unavoidable TPK).
It's not meant to represent the functional CR. The only function served by calling him CR32 instead of CR24 is to compare him alongside other CR24 dragons, and to get his data (BAB, HD, AC, Saves) away from other (non-dragon) CR24 creatures.
Although you have raised an interesting question in my mind-- I would be interested to sort the data, not by full CR, but solely by the Spine value of every critter. That would be a useful metric.
Gonna do that tonite I think.