Hi,
Ok.
"Poorly playtested" is me quoting someone. Those aren't the words I'd use, but I do agree that over the course of its run, AM5 playtesting became increasingly insufficient to handle all the new rules, subsystems, spells, virtues, flaws and their interactions.
This began to occur as early as RoP:M and TL, with baroque rules that didn't always work with other rules or sometimes even with themselves.
By the time we reached Magi of Hermes, I decided it made the most sense to treat any spell writeup as someone's house rule that happened to be printed but that also needed to be validated to the exact same extent as anything posted to these forums. For all that AM is supposed to have the most awesome magic system around, I bet that it's easier to get consensus about the level of a D&D spell than an AM spell.
And there are still arguments about how things like core MuVi and Parma really work.
Poorly playtested, maybe not. Insufficiently playtested, I think is fair. Also fair would be to note that AM5 is trying to do something different, and that there simply isn't enough manpower for decent test coverage. Which there isn't.
I also don't see any basis for AM having the most thorough playtest in the business. Certainly it cannot be in terms of playtester count, because things like D&D Next probably got more testers than there are active players of AM5. Nor can it be in terms of time per supplement, because AM5 supplements came out rather quickly. Nor can it be in terms of pre-release exposure to the community, because even train wrecks like RMU have received years of playtest and exposure, which in that case isn't going to help at all, because updating a classic game system (like AM) in a satisfying way is hard and they are failing mightily. Nor can it be in terms of "everything just works simply and smoothly, so the playtest must have been" because that's obviously not the case. An iPod, this is not.
AM5 has its upsides, to be sure. But the system is sprawling and baroque and even cantankerous, and it seems that part of the philosophy of AM5 has been to differentiate different kinds of magic by using new and different kinds of subsystems, each of which is also baroque and cantankerous all on its own, and therefore difficult to test alongside all the others.
This design philosophy is hardly unique to AM5, nor is the tension between the benefits of a single game mechanic and the blandness that usually arises from the utterly generic.
The design decisions of AM5 favoring flavor over simplicity increase the amount of testing necessary to achieve a given level of coverage, because there is so much more to test.
AM5 has good things to offer, but it also has lots and lots and lots of bugs.
Anyway,
Ken