They replied to this in the Reddit AMA and it’s a matter of combat time. Some fights in DOS2 could really drag and that was with 4 people. You have to account for summons and stuff as well. Their combat philosophy for BG3 seems to be quicker more fluid encounters, which to me personally is essential.
Well, first things first, I read what they said on reddit. This thread exists PRECISELY because they also said that it's still not set in stone and they are considering their options and checking how things go in early access. So I'd like for them to actually consider the BETTER option.
Second, the problem of "combat time" has already been addressed several times across the thread, starting with the very OP I wrote, where I pointed that other games already did a six-slots limit without much of an issue.
It's also worth stressing that just because a designer sometimes starts with notion in mind, it doesn't necessarily mean it's some holy gospel and should never been challenged.
I for one think they are absolutely wrong in worrying about the issue of "combat time" with six characters, but let's be real, what they actuall worry about it making it simpler to map on a controller.
If they could were genuinely worrying abot the overall time:
- they wouldn't implement unskippable animations that can't also be speed up in the options.
- they could try a control scheme that doesn't drive you mad when you want to to anything else than having your party in autofollow.
And this just for a start. Some competent encounter design would do the rest.
I mean, in a lot of cases a six men group would clean an encounter faster than a smaller one, anyway, not taking more time.
Oh, well, you know what? Who the hell I'm kidding, anyway?
I'm just wasting my time arguing for the sake of it. We all know they'll stick to the initial idea no matter what and if anything they'll just shrug and say "We were wrong about it" one year later as they did with their bizarre and ineffective armor system.
Sometimes I swear it feels like they have a knack for over-designing convoluted solutions at "problems" that never really existed.
The chain-unchain system in place of the usual point-and-click is a perfect example of that.
