Re: [Crawl-ref-discuss] Xom, another round

[David P. <pl...@mi...>]

On Mon, 30 Mar 2009 jp...@mi... wrote:
> Quoting David Ploog <pl...@mi...>:
>
>> I have no idea if anyone else is playing Xom.
>
> I am! :)

As Eino just proved, others do as well. That's a good feeling :)

>> By my assessment, the situation is okay for now. Not ideal, sure,
>> but good enough for release.
>
> I agree. Now that tension is factored in, Xom doesn't feel like a slot
> machine anymore.

Agreed. Though my last three test games had their own clustering: I've got 
many hostile demons summoned (but no weapon animated).

>> I still feel that the weights are out of proportion. Ultimately, we
>> may need actual numbers as weights. How hard would that be?
>
> According to the comments ...
>
>     // This series of random calls produces a poisson-looking
>     // distribution: initial hump, plus a long-ish tail.
>
> ... which probably means more to you than to me.

I understand it, after having a peek at the source. The main idea is that 
the effects on top of the if-then-else chain are much more probable (the 
hump) and thus provide the staple Xom effects. This is a good system, if a 
bit opaque -- the long tail is something really useful. In other words: 
let us not add any weights for now. (And not for later, until we've 
figured out how to combine the two parameters we aready have.)

> I've got the impression that with high severity (You are Xom's teddy
> bear) the beginning effects (might, berserk, also casting spells)
> become frightfully rare, which is a pity because they are so
> interesting effects when they happen.

Yes. In my current testing game I got a berserk while running from a 
ghost. This was really nice. The severity model is older than tension. I 
don't see a particularly good reason for the x_chance_in_y(2,sever) for 
the first effect, for example.

I don't have a consistent mental picture on how the interplay of tension 
and severity should work, however.

> In any case, the weights factor in a fixed order of events,
> abs(you.piety, 100), tension, and some negligible stuff (such as
> whether you're in the Abyss or Xom is feeling nasty), all of which
> might be tricky to reproduce. Instead we might want to focus on
> creating a new set of weights that also takes into account severity
> and tension but need not necessarily reproduce the current situation.
> However, any such change should wait until after the release, which
> has the additional advantage of garnering more feedback on how things
> actually work in the current setup.

Yes, this I agree with. Of course, Xom is a god that is never finished 
(sharing this fate with Nemelex), as the long list of cool effects shows.
In other words, that is another reason to let go for now and come back to 
Xom later. And, Xom can be fun :)

> Now that Xom's mood doesn't change as often the title announcements
> don't really bother me anymore. Still, I'd like to change the wording
> to be less technical, so instead of "Your title is now: A beloved toy
> of Xom" it should be something closer to "You are now a beloved toy of
> Xom" -- does that sound okay, or could it be worded better?

I agree.

> Also, the actual religious title is now always "Toy". We could use the
> current mood announcement messages as a replacement for that one
> (properly capitalized), and use something different for "Favour", e.g.
> "Xom is pleased with you/Xom is ambivalent about you/Xom is annoyed at
> you" or "Xom is in high spirits/feeling moody/disgruntled".

I am against such a specific description. In particular, as a pleased Xom 
can still kill you etc. Rather, I'd like to rarely have wacky titles like 
Pulsating Apple, Astonishing Handkerchief etc. [1]

> PS: Today I escaped from the Abyss for the first time in a real game,
> whoohoo! :D Xom may not have been impressed, but for me it was a very
> exciting event.

Yay! There should definitely a chance that Xom pulls you out of the Abyss 
(no matter how you get in, though the chances could be different). But 
that's for later, as well.

David

[1] Here is an observation that I made about humour in Crawl: Funny 
elements that are old are treated with high respect (e.g. the pizzas, or I 
wouldn't want to harm the death cobs). Everybody is accustomed to them, 
people grow nostalgic and so they stay.
However, adding new funny stuff happens rarely. It is hard to stand up for 
it (much harder than for weird gameplay ideas, at least in my opinion). 
For example, there was the idea that the Royal Jelly sometimes drops a 
royal jelly upon death which is simple pun, was noted as such and trashed 
as such on SF (of course, the original distinction RJ - rj is the same 
pun).
I try my best to avoid this kind of stalemate when it comes to gamplay: I 
will suggest and support removal of features as I see fit, and propose new 
things, even if considered too weird by some. As it is much harder to back 
up humour (at least for me), I won't do the same with jokes :)
Note: let's not start on the pizzas and the abundance of fruits again, 
I've made my peace with them.