Alright, I'm past most of the tutorials. I've made 5 maps and I want to improve and break the cycle.

[KineSquared]

So I've created a couple maps at this point, some of which play badly, some of which play bad, and I've made dozens of maps that never made it out of pre-alpha. Yet it seems like some issues seem to come up over and over again, and the most common complaint I've had is "awkward geometry" and confusing layouts. I've tried to address some of these in maps, though I fear I've not really made any significant improvements. In addition, 3/5 of my published maps are on a4 or less, which may just be how it goes but I keep feeling like I just build these things which require huge systemic changes.

Any advice or tips? I would love some tips, tricks, or even some kind of mentoring, as I've had little (not none, thank you yoyoyo). I'm not some complete noob who needs help setting up a spawnroom, and I would love to go over some of my old maps one on one and see if I can get some assistance picking apart the flaws in my "natural" style.

Here's a list of my works, from most recent to oldest, to help jog anyone's helpful memories if you've tested them.
Woodworks

Mintmill

Drumwalk

Astron

Polish

Thank you for any and all feedback!

 

[Wyvern]

I’ve got some bad news for you, problems such as this take time. Apologies if that sounds ambiguous, but I’m struggling with the same thing. Mapping is a very long and arduous process that takes years to perfect. Problems such as ‘awkward geometry’ and ‘confusing layouts’ all are hardships of learning, and will eventually be corrected with time. You may not nail it your first attempt, maybe not even your one hundredth, but that’s just how experience comes, through long and difficult processes that take time. You’ll eventually be able to ‘eyeball’ it as you learn gameplay theory and design, but learning a skill as such takes years, and is not completely perfectable. My point is, asking for improvement won’t make you better, repetition and learning will. Keep working, Kine, I’m excited to see what you can create.

 

[Brandini Panini]

I agree with most of what Wyvern said.

Though, finding a sort of mentor isn't a bad idea and would help with more nuanced things and basic problems like weird geometry. I think I got over certain hurdles after hearing what other mappers and comp players had to say about my maps. In the end it still will take time and learning, it really all depends on how fast you pick up on things.
I would be willing to talk to you about some stuff but I don't know if im really the most qualified, I've got quite a few maps under my belt but not enough to call myself an experienced mapper.

 

[Yrr]

keep making maps until you get all the bad ones out of your system, i can see improvement already so it tracks that if you keep doing it you'll just keep getting better

 

[B!scuit]

I'm surprised how well i remember these maps just from the screenshots
I really like the amount of height variation & overall pacing / density in Astron & Drumwalk
Those 2 are the ones I remember & from what I remember they have really well separated play spaces while still having layouts simple enough to easily follow

While at times they can feel like a series of filled boxes, imo they play well and are easy to follow on your first play, which I think quite a few official maps struggle with

You're doing a lot better than me at all of those things I just mentioned, the things you're not as good at you really have to identify for yourself
Once you have identified something you want to improve, finding a specific goal like: "I want to understand chokes better" is miles better than being unsure of your own ability and only asking: "What am I doing wrong here?"
I've been taking this approach of having a goal to a map (i.e. use lots of open areas) as a goal over just simply tweaking to fit feedback
I'm not sure how you approach designing a new map but I generally go for visuals first, which always hurts gameplay.
The goal-driven approach definitely seems to be resulting in better playing maps

If you really want some hard feedback on an older map pull them out and get them tested again
There's a few little tricks like using non-dev textures to get testers to focus their feedback on a certain spot and I've found if you want good feedback you need to create something that players can focus on
The closer to something you'd release the stronger & more perfectionist feedback can get (at least, I've found pretty grey alphas draw more feedback than dirt & wood alphas)

 

[RaVaGe]

Been a while since I posted something here.

First, the most basic thing is to understand the gameplay concepts behind every gamemodes and the game itself, if you don't fully understand how a map works, in terms of flow and guidance, then you have learn these concepts right now, because you can't master something you don't understand. Understand the pacing of the maps, where are the bottlenecks/chokepoints and why they are there, look at valve's best maps and ask yourself "what is making this map such a great map".

Secondly, you have to make the distinction between level design and art. In the game industry it's a totally different part of the job, there are level designer and there are artists. The level designer create a fun experience for the players with a great flow and design, while the artists are working on the existant layout and create an immersive experience.

Thirdly, understand how optimisation works, if you can't optimise your map on your first alpha, you're dead.

So, as a newbie, I would advise you to create great levels first, then work on small details projects on the side, make some detailled barns, bridges or whatever, but as a side project. Once you understand how to detail, you can scale this up to a full map.

When you create a level from scratch you have to understand that it will sucks, the whole point there is to make sure the changes you are going to make are as easy and fast as possible, iteration is the key in the first days of your map, test and tweak, don't listen too much to the feedback, and look how the map plays, then make the changes to what you believe will make an improvement to the gameplay.

Complexity is often a side effect when you are starting, most aspiring LD believe that more is better, it's the exact opposite, less is more.

Work on a 64 unit grid, use two dev textures, one for the floor, one for the rest, add another one for the displacements .

Create some elements that you are going to use for every elements of your map, doorframes, windows etc, you can even use instances, make sure to use func_details for them.

Don't use any props/decals/lights during the first version of your map, props are there to improve the visuals of your map and to catch the eyes of the players, consider them as bonuses to improve your map when nothing else has worked. A good article about that : http://www.nodraw.net/2010/08/tf2-density-of-detailing/

As for the lighting, make it as flat as possible, you don't want to interfere with the natural flow of your map yet. Check this out : http://magnarj.net/article_funclight.html

Here's the spawn area in koth_brazil during the first alphas :

Then I started to break it down

You can check some of the other maps i've made, they follow the same patterns, I consider 3dnrav as the pinnacle of what I could achieve in terms of simplicity during the first alphas.

https://tf2maps.net/threads/3dnrav.26724/

I could go on and on, but the point is, simplicity. Understand the core of what you want to achieve and work around that in the simplest way possible.

Good luck, it took me more than 10 years to achieve something relatively good on hammer, patience is the key

 

[Idolon]

First thing I'd advise is making smaller maps. A standard 4cp payload map is harder to make work than a shorter 2cp payload map because there's just a lot more stuff. I'm still struggling to make 4cp work.

Second thing I'd advise is placing more restrictions on your layouts. Something I can pick out from your images (and is probably the source of the "awkward geometry" comment) is that your layouts seem to meander a bit. There's a lot of corners, lots of turns. A lot of people do this, and it makes layouts difficult to reconcile.

Valve maps tend to be very simple in their overall layout, and it generally helps the layout make more sense. Look at Goldrush/Dustbowl stages 1 and 2 and you'll see that the overall "gesture" of the layout only makes one turn. The overall gesture of Upward's layout makes 4 turns. Gorge and Granary are straight lines.

The last thing I'd advise is trying to get away from putting little walls in the middle of gameplay spaces. A short wall or prop on a ledge to create cover is fine, but looking at Polish's mid layout... it looks bad and will likely play poorly. Blocking sightlines is, of course, important, but it's best to try and do your major sightline blocking work with larger and fewer elements - like how Badlands mid controls its sightlines almost entirely with just two train cars, or how Granary mid controls its sightlines with four stacks of containers, or how Badwater A controls its sightlines with rock props and a giant hill.