[@Shoryu Magami] When I say my chars might see something, I mean they might put pieces together that I would not.  For example, I immediately went "ooh maybe ghosts?!" when you said someone might stick around after dying.  Another person might have gone "right, so...rolling new characters?  Okay then."  And a third might not even have noticed.  It's an issue of the player having a way of viewing the world that does not always match the way their characters think, and yeah, there's only so much you can do about it.  My social ineptness bleeds into my characters no matter what I do, even the ones that are supposedly very perceptive or very compassionate.  ...This can and has resulted in them being total jerks at times without me intending them to be.  I do tend to pick up certain things better in text, simply because of conservation of detail.  This means that the more details you give me, the less likely I am to notice the important ones, in many instances.  And of course if I'm the sort of person you have to throw a brick at before I recognize a guy hitting on me, many of my characters are going to be similarly dense without a bit of serious hinting or OOC help.

As an extension, I have been known to ask "Is character A acting X" where X is nervous, upset, or some other emotion.  Or I'll ask "would [my char] be able to recognize that A is X", with X again being an emotion or some such.  There's also the matter of someone trained in fighting, for example, might recognize a soldier by their behavior, but some office worker might not.  If you say "A cleaned the mess with the conservation of motion expected from a martial artist", I'm going to take that as it's obvious enough that there's a good chance even the office worker would notice.  Not for /sure/ -- I still get to take his nature into account and decide if I agree he would.  But it's a possibility.  Thus too /many/ details can actually create issues, because I highly doubt you describe every detail and then also just how evident each detail is to various types of people.  I'm not sure anyone would want to read something where you have to hold hands like that.

It helps that I draw, and I do have a slightly anime style in many cases.  However even with anime, there's a wide range of styles.  You might be better off creating tokens with names, or descriptors -- "shadowy figure", "glowing eyes", "thin lady".  That way you can preserve consistency between tokens, and still use the map.  It would require more reading, but still work.  As for anime-ish, if you mean in imagery, that's not the text at all and so it would limit portraits /only/ to anime ones, but not require portraits for everyone.  If you mean in style of story...that's not an image issue, I've seen it done in books and TV shows, and even rps that used varied images or no images at all, and so there is no reason to constrain images because of it.  Personal opinion here, though.

As for the overall plot in opposition to the game of chupa...that's not how you presented it, and you're going to be attracting people who want the latter, and will gun for it, and make characters with that sort of a mindset instead of a "we probably should be working together, guys" kind of mindset.  This will affect your outcome.  It will affect the players going in knowing this is not, in fact, a game of chupa.  Also I'm fairly certain I know not only the tag but also the twist, and possibly even why.  And with me knowing, I'm more likely to have my character put pieces together faster.  I'd need evidence, of course, to imply such a conclusion, but it's a lot easier to put a puzzle together when you're looking at the box.  I'm not sure this is fair to others, and so I'm going to bow out.  Between my uncertainty and the rest, I don't think I'd make a good player.

Totally gonna lurk to see if I'm right, though!  <3