One missile, doing 1d6 + Int (or Cha) mod damage.
However, what I would really envisage is something like Arcana Unearthed/Evolved, where each spell presented a Heightened (and Diminished) version in the spell text. So, you can get two (or three) missiles if you want - just prepare it as a 2nd level spell.
Doesn't work. If some spells increase with level and some do not, you might as well just remove the ones that do not - they're obselete before they've even been published.
Even as things stand, the optimisers have identified that some spells do 1dX plus the caster's level, while others do a number of dice depending on level - and that spells in that first category aren't worth taking.
Part of my fix was to give casters more spells at low level, which should reduce the likelihood of this. But if the caster truly is out of spells then they really
should suffer - that's the trade-off for those spells being awesome.
(Oh, one other thing - amongst the various metamagic options, I would suggest options to prepare a spell for use on an At-Will or Per-Encounter basis. So, if the player of the Wizard
really doesn't want to risk running out of spells he has that option - with the trade-off that he would have to stick to weaker spells.)
The problem here is that, as Hussar pointed out to me on another thread, beyond very low levels scrolls represent an entirely trivial expenditure - using WbL, a mid-level Wizard can
easily afford about 100 low-level scrolls, effectively bypassing Vancian casting (and, in the process, effectively rendering the Rogue's skills useless).
So if we're going to fix the issue by "making scrolls cost a lot more", that
actually means multiplying the cost by a factor of
at least 10.
(Incidentally, the rarity isn't a problem. In general, allowing the Wizard to use any spell
once isn't a problem (even if it's a much higher-level spell than he could normally cast). The problem comes about when the Wizard can craft/buy
multiple uses of spells.)
Agreed.
And agreed.
Agreed, except for the very last. I would actually present an option of three different casting mechanics: Prepared (like the Wizard), Spontaneous (like the Sorcerer), or Mana-based (or whatever; works like the Psion). The first time a character takes a caster class he chooses one of these options.
In each case, the caster trades off long-term flexibility (max spells known) against short-term. That is, the Wizard can know many spells, but only prepare a few per day (with no substitutions later); the Psion knows very few, but can cast them in any order provided he has the power points available. (The Sorcerer, of course, lies somewhere in the middle.)
But I agree - no caster should simply know all the spells for his class. That made sense in BD&D, and perhaps even in 1st Ed, when there were only a handful of spells; when there are hundreds of spells at each level it's pretty crazy. (Plus, that immediately gets rid of the whole "all Clerics look the same" problem.)