EzekielRaiden
Follower of the Way
This is a perennial paradox of 3e-like design, you see it in 3.x, PF1e, and now 5.x. Martial characters are much, much easier to play "straight," but almost always really hard to optimize once you have a solid background in the mechanics, because the resources are so much more scarce and the features rarely play nice with one another. Conversely, full spellcasters (or near-equivalents, like the 5e Warlock) are all very complex to play at baseline, but once you know your charop stuff, the world is your oyster, they've got so many features that you can play with, and most of them interconnect nicely.I think any of these classes can be complex in building beyond the basics, especially when adding in multi-classing, but what happens when we multi-class the Champion in crit-fishing builds? Seems very complex to optimize.
I voted based on my impression of average play, not double thinking build optimization.
Personally, I prefer to judge complexity based on baseline, rather than on pure charop potential. There was actually a vaguely-similar argument WRT 4e classes, albeit without being rooted in caster/martial disparity. Warlocks had much lower output when not optimized compared to (say) Sorcerer or Ranger, but if you did optimize they were just as good as any other option. The "skill floor"--meaning, the minimum skill required to play effectively--was higher for Warlock than pretty much any other Striker. But the "skill ceiling"--the point where diminishing returns kick in for player skill--was pretty much equal to any other Striker.