Again, RC, the issue at hand isn't the idea of having signs of a monster's presence available.
I think that depends very much on who is "speaking". It may not be
your issue at hand, but some of the posts on this thread make it seem to be.
Nor is it my argument that "there are never situations where PCs don't come across those signs or find themselves without the knowledge to interpret them correctly" -- that is Hussar's interpretation of my argument.
I merely require that it be possible. Whether or not the PCs actually come across those signs, and whether or not the players interpret them correctly, is besides the point. I only require a "footprint" that is "consistent and self-referential".
Such as the comment about Maenar leaving vermin bones behind. Plenty of things eat vermin - does the expectation of the PCs identifying the Maenar rely upon them having a ranger with an incredible Nature skill?
You misunderstand -- is the maedar of the "all flesh must be eaten" variety, or are there some things not worth turning back to flesh to eat? If the medusa encounters a giant spider, does the consort turn it to flesh to be eaten? Or are there some things that are simply left stone?
If we are to assume that the medusa is smart, she would surely know that a stone spider attracts fewer wandering monsters than a rotting flesh spider. And, assuming that one is using the wandering monster tables provided, she would know that some wandering encounters may well cause her harm.
IOW, PCs wander down to Level 7 perhaps once in her lifetime (

) but she has to worry about wandering monsters down there 24/7. What makes more sense? Leaving rotting piles of uneaten flesh around, or allowing some statues to remain statues?
Similarly, is it that unreasonable to assume that some encounters may happen without tons of advance warning?
Again, you seem to be reading too much of Hussar's "What you are saying is...." and not enough of what is actually being said.
My argument is, basically:
If a game world is to have internal consistency, all creatures that can be encountered should have a "footprint" which is consistent with their effects on the environment. The greater the environmental impact, the greater the chance that the PCs will encounter some element of that footprint. They may or may not actually do so, and they may or may not have the knowledge to determine what the footprint actually indicates.
Some may disagree with this, but I note that even Hussar's arguments are not that the medusa wouldn't have a "footprint" per se, but more about how she might be able to reduce the opportunity to discover that footprint and/or interpret it correctly.
As you say, there can be all kinds of reasons that the PCs don't encounter warning signs prior to an encounter, or misinterpret what signs they do encounter. After all, warning signs are part of the
environment, not part of the
encounter.
It is only when one begins to think of GMing as making a series of discrete
encounters, as opposed to the creation of a continuous
environment in which various encounters may (or may not) take place, that these problems arise. IMHO, and IME.
Nowhere have I said that the PCs must stumble upon a creature's footprint in advance. Each time that others have tried to claim that this is my position, I have attempted to clarify that this is not my position. Indeed, the player's knowledge that they might have missed/misinterpreted clues is vital for the game to remain interesting, IMHO, and as I said upthread.
Take it away from SoD creatures, and the same point is vaild to exactly the same degree.
Example 1: The game includes powerful monsters, which are capable of handing low-level PCs their tushies quite easily. You can (1) make a world in which it is impossible to encounter such creatures at low levels, (2) allow the PCs to simply randomly encounter such creatures at any level, or (3) create a self-referential world in which it is possible to learn that spiders dwell in Mirkwood, goblins in the Misty Mountains, and Smaug in the Lonely Mountain, and let the players choose what they feel up to dealing with.
They might discover trolls, and stone giants, and eagles, and wargs, and Beorn, and elves, and Gollum as well.....that choice doesn't mean that there is nothing to learn. The people in Laketown might believe that the dragon is a myth, and Gandalf might tell them to take a road that goes nowhere (good thing they left the path after all!) -- not all advice is good. Even Elrond was unaware that the goblins had opened a door on the "safe" pass.
But all of these things could, potentially, be learned prior to being met. And, for many of them, Tolkein makes certain to tell the reader what "footprints" they had (even if the dwarves and Bilbo never learn of it). The trolls are stealing sheep from local farms (or worse), and Elrond's folk know that trolls have come down from the mountains (and warn Gandalf). The goblins know something foul is down by the lake, even if they never impart that to Bilbo before he meets Gollum. The goblins are also well aware of the great eagles of the Misty Mountains....as is Gandalf, who once rendered their king a service. Etc., etc.
Example 2: The Wolf-in-Sheep's-Clothing.
This monster (originally from
Expedition to the Barrier Peaks) almost never works for DMs, because most DMs forget the necessity of providing elements of the natural world with a viable "footprint". Bunnies are never encountered, ergo, if bunnies are encountered now, they must not really be bunnies.
Similarly, the DM never mentions ravens normally, ergo this must be a wizard's familiar.
----
Providing "footprints" is, IMHO, just part of good GMing. If you do it well, players will engage with the world you offer because it is interesting and profitable to do so. Fail to do it, and there is little wonder that things just seem to pop out of the woodwork, or that the players never mistake that Guardian Familiar for a normal cat.
------
EDIT: On further consideration, the problem Hussar describes with SoD monsters is rather similar to that he has described for so-called "gotcha!" monsters in the past. If one assumes that a creature exists merely to appear in a discrete encounter, no wonder it seems to be some form of "gotcha!" When we discussed the rust monster, for example, I was able to describe uses I've made of that creature within actual play, which he seemed to believe required some unusual convolution of thought, whereas I considered that normative for what I call "good GMing".
Perhaps this is an artifact of following the DMing advice in the 3.x books, where the "unit" of an adventure is an encounter. I recommend instead consideration of the whole, and then, when writing individual encounters, trying to imagine how what is written will affect the whole. Do the inhabitants of the ruins know that there are intelligent otyughs living in the Filthfall Middens? If so, do they talk to them? Are the Middens a neutral truce ground, because everyone needs to dump their trash, or are they the site of constant skirmishing? In either case, what do the otyughs know as a result? Etc., etc.
When placing wandering encounters on a table, it makes sense for the GM to consider what effects those encounters have on the environment. A dungeon level with wandering gelatinous cubes looks very different than one without. Wandering rats will leave rat droppings (which become part of the dungeon dressing). Rust monsters consume the hinges from doors, knobs, pull rings, and so on. Again, etc. etc.
RC