When did I say he'd maintain a career like Rivera's? We have him for three years, not 13. And disingenuous? I am deliberately trying to deceive people? Miller's best ERA+ is 198. Kimbrel is coming off three years from 228-399 prior to 2015, when, as discussed above, he got off to a poor start (BABIP fueled) and was dominant from June onward. You can disagree with my reading of the stats but I don't know why you're questioning my motives.
I apologize for my poor word choice, I didn't mean to imply that you had hidden motives; merely that the statement itself is disingenuous. The stats you've painted so far are misleading for several reasons.
1) ERA+ is problematic because the relationship between ERA+ and actual ERA is skewed as ERA trends towards zero, because ERA+ trends towards infinity. These big numbers may seem daunting but are not very different from one another. For example, the difference between an ERA+ of 228 and 198 is 0.29 in ERA. Over 40-60 innings, this amounts to 1-2 runs which is smaller than year to year variation. The 200 point difference in ERA+ is 0.9 in ERA. Over 40-60 innings, you're talking about 3-6 runs, which is slightly larger than year to year variation. This may seem like a big difference here, but we've got an SSS problem.
2) ERA is not very stable at 40-60 innings, regardless of how one evaluates the amount error in the measure (e.g. via correlation, or resampling approaches like bootstrapping). Although the observed values certainly differ between miller and kimbrel, that difference is more likely to be much smaller if we played out the same season 100 times over; in fact, in many of those 100 seasons, miller will probably have the better ERA. This isn't to say that miller is better than kimbrel, but rather that 40-60 innings is a noisy sample. Peripherals are probably better indicators if one relies on single season reliever data, but 40-60 innings is pretty small even for that.
3) Ignoring 2015 is ignoring the most recent data point, which is important data that should be included. The fact that he got off to a poor start is data that shouldn't be ignored just because it doesn't fit your hypothesis.
4) Miller has been a reliever for four years. Those four years provide a sample size of 200+ innings. Even including Kimbrel's outlier 2012 year, the difference in four year ERA/FIP/XFIP is 0.77/0.5/0.34. This is not a huge difference, and almost all of it is driven by 2012. Throw out 2012 and the difference drops to 0.22/-0.05/-0.26.
A statement like "Miller's best ERA doesn't match any of Kimbrel's years prior to 2015" implies that Miller hasn't even come close to Kimbrel's level. As shown above, such a statement obscures a proper interpretation of the data.
EDIT: Finally, let's assume that your comparison to Mariano is accurate; Kimbrel is an unusual commodity in that he's a low volatile dominant reliever that won't have a bad year. If this is how the market views him, then you'd only need to add one or two pieces (e.g. devin/marrero) to pick up a top-line starter. Furthermore, you wouldn't need to put together a package for yet another SP (unless you decide to outbid everyone for Price despite picking up 10 million in AAV).
