The temptation overcomes us all 😁

Yeah, same. This video gave me a fresh reminder to move back to this blocking mode. I started to cheat myself out of it. 😅

Yep! :D It's all about finding a balance in the process and focusing on what's important at which stage. When you've gone through several iterations and have a solid-playing layout, it often can't hurt to spend time on selling the visual intention (though in moderation!)

I can tell you from experience that being there early when this problem happens can be a tad frustrating when you can do very little to change it 😅

Your early blockouts don't need to be visually detailed, but the rough composition can (and often should) visually represent what you're building, to some extent. You can colour code to indicate what are rocks, water, sand, buildings, etc., but keep things very simple! Your early versions simply aren't going to be pretty, and this is fine. That's not to say you can't have extensive research and explorations in the setting, theme and even visual detail tests parralel to this, to make sure those aspects are still an integral part of your design instead of an afterthought. In the industry, concept art to support early blockouts, many syncs between designers and level artists and constant iteration on both design and art are crucial iterative loops to make sure gameplay and visuals go hand in hand.

If you're making something like a CS map on your own and are relying on existing assets, I suppose you'll have to use some of the actual assets from the start because they form one of the core restrictions in what you can build. In that case though, keep it to a bare minimum use of the larger, structural assets you need to shape your level and avoid all the small stuff as long as you can :)

Mostly through a ton of research and reference + iteration. The more you practice and learn, the more some things become intuitive. But for realism, look at the real world around you for reference too :) Whether it's architecture, or the way a path winds around natural obstacles and follows terrain in the forest.

ok thanks ! ;)@@timdoesleveldesign

In larger companies, when a blockout level plays well, usually there's some sort of "greenlight/validation" stage. If approved, level artists will start replacing the blockout with actual assets/props. You could follow a similar process, iterating on the blockout until you're happy with the way it plays. Then more or less build "on top of" the whitebox, bit by bit swapping parts out, focusing on the larger, structural stuff first before worrying about the detailed dressing.

And keep in mind that no matter how good you are, the norm is that you never get things perfect in the first pass! It's about iterating step by step, testing, and critically assessing what works and what doesn't :)