Matter of choice and personal taste, I would say. 

For big ticket items, i like to have an obvious cue saying "i am a secret" like a megahealth behind bars, or a quad on a high ledge or something, to tease the player. This challenges people to figure out how to get the item.

For any secret, even a +25 health, i think it's important to have a subtle clue that tells players how to get it -- slightly rotated texture on a shootable door, or one light fixture that is a different color than the others, or whatever. The cue can be MORE subtle in cases where the tease is also present, becuase then people will be trying stuff anyway. (shooting walls, exploring, etc.)

As an alternative to the subtle cue, i like to put the access to secrets in a place that players often don't think to explore. This can be becuase it's obviously not a route, like the pool of water in house of desolution. Or, it can exploit the fact that players don't think about how moving objects change the environment -- the quad under the lift in contract revoked is a good example. You can also exploit the player's limited spatial sense, especially with regard to the vertical dimention.

Or... well, i'd give more examples, but i don't want to give away all the secrets in my unreleased levels :) 

Or... well, i'd give more examples, but i don't want to give away all the secrets in my unreleased levels :)

Heh, too true.

While cramming secrets into Contract, I tried to go for variety.
Some of the secrets are intended to have a direct benefit on subsequent combat, the lightning gun hidden in the first map is a good example - the LG is the best weapon to use against shamblers so providing it at that point allowed about 6 shamblers to be defeated more easily if the player found the secret. For that reason, the 'tantalizing visibility' trick was used so the player would at least be aware there was something to find.
For more powerful, potentially devastating benefits like a pentagram, I have no qualms about being an utter bastard, tho I always endevour to make the secret an integral part of its surroundings, like using the chequered platforms or wooden beams.
I guess I assume that any player will try to explore every reachable area of a map even if it's not part of required navigation; especially if it's not part of required navigation. I do that in every FPS I play - I'm a nosy cunt - so I tend to put secrets in corners, dead ends, pools , pits and so on.

In terms of normal navigation, I always try to make the path explicit. Arrows I find are a hokey way of directing the player because they look incongruously generous for an environment that is otherwise trying to kill the player; a last resort only. One of the important things about SP is giving the player room to look around - places to stand, distance to jump - so they can gradually build up an impression of the overall construction of the map. It also helps if the layout is based on consistent elements that the player gets used to.
Lighting is also critical - ffs, don't leave the most important exit in a room in total shadow. That can be really frustrating. Weapons should also be delivered to the player in a way that is unmissable even to a myopic elephant.
Basically, the more important a structure/item/route is to progression, the more prominent a place it should take in the player's perception. 

another rule i try to follow is: never give the player a key until after he sees the door that it will unlock. This is a specific case of a more general principle, that it's good to have places/things that a player sees but can't get to until later, creating anticipation/desire and then later providing closure. It also defeats the purpose of a key to have it unlock a door that you don't get to until you have the key, becuase the door never blocks your progress, then. 

same goes for switches really; see moving part -> find switch.
Also, if a switch is remote from its target, there should be something to help the player match up action with consequence: center print message, texture similarity etc. Mostly important in a map with lots of doors or switches. e.g. all the shootable switches in Contract were for secrets only; all the same animating texture, different texture from 'normal' switches.
Consistency, people, consistency... :P 

I liked all the custom Q2 maps that used the destructable-crate texture on non-destructable crates. 

I think metlslime is pretty much bang on.

The are two things I find I say to myself which signify a bad map design. The first is: "so what the hell do I do now?" And the second is: "why the hell did that happen?" Saying hell is an integral part of it. :)

Do what Nintendo does, not what Cyan Worlds does. 

I liked all the custom Q2 maps that used the destructable-crate texture on non-destructable crates.

On a somewhat related note: is there ever a point when these design elements become cliche? And if so, how wise is it to try and change it to something new? 

they can become cliché. But the principles behind them are universal, i think. 

cliché? 

clich� 

i typed clich� both times... i wonder if my change did anything. I can't see how, as all i did was add calls to PHP's addslashes function. 

Well the little bit above the E looks a bit like an apostrophe. Maybe the board has fuzzy logic, like those new washing machines. 

Fuzzy logic: like wooly thinking, only less so 

now the two that were broken (posts 9 and 10) are fixed, and the "fixed" ones (posts 11 and 12) are broken in a different way. 

they can become cliché. But the principles behind them are universal, i think.

yeah, but the issue I kind of want to bring up is the communicating to the player those underlying principles.

ex: let's say I thought that using the gold key to open a door was cliche. So instead, I make it so you have to kill a certain monster to open the door. So now how do you know you must kill the monster to open the door?

So, I guess, I'm wondering if there are any kind of, I don't know, universal concepts or important ideas to know while trying to convey to the player what he can, or needs to do.

For instance, I like kell's comment on making sure the important exits are well lit.

The are two things I find I say to myself which signify a bad map design. The first is: "so what the hell do I do now?" And the second is: "why the hell did that happen?"

yeah, personally I agree. But now I'm interested in taking it one step further and asking: what helps make maps understandable, or not?

I don't know, maybe this is an intuitive issue, which can't readily be discussed and analyzed. Players either understand your map, or they don't. Anyway, I just thought I'd bring it up. 

"ex: let's say I thought that using the gold key to open a door was cliche. So instead, I make it so you have to kill a certain monster to open the door. So now how do you know you must kill the monster to open the door?"

Get the map going early on with a grunt almost right in front of a door. Every one kills the grunt, and everyone will be watching when the door opens. They'll easily make the mental model you want and now they know that it's in your bag of tricks. And now you won't have to resort to printing "X more to go" messages on the screen, which is a terrible kuldge.

It's like in Zelda: everyone knows the defeating all enemies, pulling switches, lighting torches, etc are possible tricks. Thus puzzles which might be too hard in other games are more reasonable in a Zelda game. It's all about helping the gamer form the correct mental model. 

set a precedent.

A good example is in rubicon, where i have ladders which are really just really steep staircases. At the time (1997,) those sorts of ladders were not common, so i made it so you couldn't even leave the first room of the level until you figured out how to climb. That way, later on, when you see a ladder in a combat situation, you won't get killed trying to figure it out. 

...you DA MAN.

P.S. I spent an hour and a half trying to get out of that first room, shooting and pressing everything in sight, I even no-clipped to see if there was a secret exit I was missing. Eventually I resort to imp 9 and rocket jumped out....although the weapons did make the rest of the map rather easy... 

i really AM da man. 

I walked towards the ladder..and shot up through the hole.
"Oh", I thought, "that'll be a really steep staircase then" and carrried on into the water.

:P 

any more tips? any "tasteful/subtle" solutions to not even being able to understand the precedent? have an Monster/NPC run up the ladder?


and just to add to the mix...
http://celephais.net/board/view_thread.php?id=3226&start=15

One of the best things about this map: instantly understandable layout.

my obvious question: Why? 

Of course bloody not!! 

but somehow, Shamb, I can't imagine you completed Rubicon without impulse 9

:P 

"One of the best things about this map: instantly understandable layout.

my obvious question: Why?"


It would be interesting to see a study done on what makes map layouts effective but here are some general reasons why lun3dm4's layout is so easily understandable:

1. Open Space. It's much easier to pick up on a layout when the player has plenty of space and subsequently a clear field of vision. Extremely tight maze-like maps hinder this and are harder to learn.

2. Structure Variance. Every single area in lun3dm4 is uniquely built, whether it's the main atrium with the jump pad, the pit-like megahealth area, or the carved out cliff area by the red armor. The player can tell exactly where he or she is at all times because of this.

3. Well-lit. It always helps when a map is nice and bright.

Perhaps this helps.