Build a character. Play that character for one session. Then give the character to someone else. Have them play the character for one session. Now retire the character, and give him to the DM to play as an NPC. Was the character mechanically different in those sessions? Do his abilities change?
If a Vancian spellcaster, almost certainly - the spell load-out will reflect the priorities of each player/GM.
[MENTION=73683]Dannager[/MENTION] has elaborated on this point upthread, and also made the important observation that the
impact of a given spell load-out will differ depending on whether a player or the GM is deploying the character.
Every D&D wizard I've ever seen written (2e and 3e) has been a member of the wizard class, has had levels in that class, and has learned, memorized, and cast spells in the same way that a PC wizard would.
In AD&D I've often seen NPC wizards with randomly detemined spell books and randomly determined spell load-out - this is the default in classic D&D.
Whereas most PC wizards don't have randomly determined spell books - they hunt out or research the spells they want - and they certainly don't have randomly determined memorisation.
Since there is no skill for that, I don't see your point. The number on a sheet is what determines what is possible of a character in play. Why else would we spend so much time on those numbers?
The number on a sheet is one determinant of what is possible for a character in play. The resolution mechanics are another, equally or even more significant, determinant of what is possible for a character in play.
Rolemaster and HARP rank skill bonuses from -25 to +100 or higher (there in in principle no upper limit, although going above +150 is pretty unusual). And it is not uncommon to have party members where some are at or about +0 in a skill and another at +100. Resolution is on d%, with typical DCs set between 70 and 200. The gap is therefore about equal to the die size, and not as big as the typical DC spread.
In my 20th level 4e party, on the other hand, the biggest skill gap is in History: the fighter has +9 (+10 level, -1 stat) whereas the invoker has +29 (+10 level, +5 training, +5 item, +2 race, +2 familiar, +5 stat; I'm leaving out the possibility of a further +1d8 buff from Memory of a Thousand Lifetimes). Typical DCs are between 18 and 34. So the gap is about equal to the die size, and bigger than the typical DC spread.
This might make it seem that the 4e fighter has a better chance at hard History stuff than the RM PC with a +0 history bonus. But I'm not at all sure that's true, because really challenging stuff in 4e is going to occur in a skill challenge, not a single check, which requires multiple successes. The invoker uses History (and other knowledge schools) as his principle way of engaging the gameworld. The fighter, on the other hand, is a far more physical character. And the resolution mechanics reinforce that pretty strongly.