The only competition that should be considered is the one that the most people actually want to participate in. What should and should be (for any satisfaction of fairness or propriety requirements) is completely irrelevant, unless the majority / plurality wants the most fair / proper competition possible.
Opposed checks (preferably simple opposed d20 rolls without modifiers) could be a good way to determine these factors when there's no fair way to determine them otherwise.
As I said, you run the matches. If/when it becomes clear that initiative is why one side or the other wins, then you do math. If it's not clear that initiative is why one side wins, then math doesn't come into play.
Honestly, rather than having all of the preferences for this and that weighted, I think it's better just to have a few random maps and randomly determine the starting locations and distances between characters within them. And there should be no prep, or very little. Like maybe a caster is allowed to pre-cast one personal range spell (like mage armor, though it could be whatever you want) with a casting time of 10 minutes or less as the equivalent of being able to put on armor.
That way you can see how well a character does overall, because you aren't always going to end up in the ideal circumstances for your character.
Pretty easily.. How?
But then you would theoretically need to play through every scenario a few times for fairness. That's a lot of matches.
Subtle Counterspell.Time Stop is the main one that comes to mind.
I think the next step is a formal definition of the rules and win conditions.
One map will have VERY little to hide behind. Other maps will have a decent to good amount.Are there things to hide behind for stealth classes?
As sneaky Dex classes like rogue, ranger, bard, and monk can play differently if they can hide at the start.
Honestly, rather than having all of the preferences for this and that weighted, I think it's better just to have a few random maps and randomly determine the starting locations and distances between characters within them. And there should be no prep, or very little. Like maybe a caster is allowed to pre-cast one personal range spell (like mage armor, though it could be whatever you want) with a casting time of 10 minutes or less as the equivalent of being able to put on armor.
That way you can see how well a character does overall, because you aren't always going to end up in the ideal circumstances for your character.

(Dungeons & Dragons)
Rulebook featuring "high magic" options, including a host of new spells.