First of all, let me welcome you to the Firehawk Games Forums! I hope you enjoy playing Novus for many years to come!
The Familiar Spirits table (Novus Deluxe p. 62) lists the familiar in alphabetical order. Given that it's a 2d10 roll, the results are biased towards the names that lie in the middle of the alphabet (Goat, Hound, Lizard). Is this a holdover from the original d20 roll?
Original d20 roll? This spell was not based on anything from a d20 system. This spell was actually inspired by a similar spell from a system called "The Arcanum" (i.e. The Arcanum was a huge inspiration and influence on Novus) and the spell in that system actually used a percentile roll, but it also had a lot fewer results to choose from as well.
The list of potential familiars in Novus actually changed several times over the course of the Open Beta, evolving as the system (and the critters in the Creatures section) evolved. And I pretty much just made the list alphabetical since that was easier to make sure I didn't miss anything.
Yes, 2d10 does have a bit of bell curve for results toward the middle, but don't forget that you can use Boon Points to adjust this roll by up to 2 points in any direction per Boon Point (and you can even expend Fate Points to gain Boon Points for this roll). And let us not forget that Fate points can be used to add extra dice or bonuses to rolls as well, thus increasing the chance of earning Boon Points when casting the spell as well.
Also, the caster may dismiss the Familiar at any time, and then recast the spell (so long as it has been a year or more since the last casting). So, this also means that as the caster increases in skill and power (i.e. more ranks in his Spellcasting skill), his chances of getting Boon Points naturally also increases..
So while there is a little randomness, there is also that element of the player having some control over the results as well.
And in looking at the spell this morning, it just occurred to me that I could have added a Casting Option (+3 TN/+1 SP) that allowed a for a +/-1 to the roll that determined the Familiar each time it was taken.
As for which Familiars are more common, well, there actually isn't any that is more common than any other, and the GM is quite free to mix up the list or change the chances of acquiring specific types of Familiars based on his setting. Actually, I would encourage doing so!!
For example, if there is a clerical order that worships a spider-y god, and her influence on the "web of life", then for members of that order, I would remove ALL Familiars but Spiders, as that would be the only one that makes sense for that clerical order.
Would it make sense to place more common familiars types in the middle of the curve? Alternately, a percentile roll would make the results linear (although I can't find any other cases of percentile rolls in Novus).
There are no percentage rolls in Novus. That was intentional.
