TL;DR: Everybody's telling me the solution to my problems is X, only X isn't in the game. So what kind of crappy solution is that?!
Longer version:
Go read Angry GM's take on resting. It's fairly typical and several posters give me the identical spiel, often complete with the dismissive tone:
http://theangrygm.com/ask-angry-resting-in-5e-and-why-its-fine/
The spiel, btw, is "just add time constraints through your story". But that's just dishonest - it's not part of either the rules or its supplements. The game should IMNSHO work right out of the box, and story-driven constraints on resting frequency simply isn't part of the game as shipped. Like
at all.
Yes, I've also noticed the same problem. And I've also
always gotten a condescending, "Just add time constraints," response over and over and over again. It's particularly infuriating when you tell people you're planning on running a player-directed sandbox game and the response is, "Well sandbox games can have time constraints, too." No, you twit.
We want to discourage certain resting patterns without railroading the players. This is like saying, "Downtime activities are broken, so I never give the players any downtime."
How do you make attrition work in a game where you don't fancy doing all the hard work, and instead rely on official published supplements?
How many encounters and short rests do you have per long rest? What does the party need to do when they feel they need to stop and rest? What's stopping them from doing this?
I should state out right that at low levels, the game works alright and there isn't much of a problem. Below level four or seven (or so), heroes are certainly so fragile no combat is truly "trivial" and they will feel the "attrition" even before they've used up any resources!
Feel free to use existing modules as examples; anything from Rise of Tiamat through Storm King's Thunder and Tales of the Yawning Portal. Just keep in mind that I'm mainly thinking of the mid game and above. If you absolutely must have a specific level to discuss let's use level 10; that way every published campaign qualifies.
The only constraint I'm asking of you is that you can't dismiss or "solve" the issue by the flippant "just add time constraints to the adventure" thing. Trust me, I've been given that piece of useless advice enough times already. I am specifically asking about ways on how to make D&D and its rules work, given the assumptions that 5th edition suddenly places upon the game.
The problem as we identified it wasn't one of lack of attrition. It was just one of combat encounter difficulty. That's really the same thing, just taken from a different point of view. What have we done to combat the issue? Cranked the average combat encounter difficulty to Hard, Deadly or higher as levels increase. Instead of slowly cutting the PCs down with a thousand cuts every day, you chop them down very quickly by making combat encounters consume more resources. The PCs either rest or die, and in some cases you have to pick your battles very carefully. To us, that's very much like 1e AD&D, where combat was what everybody was trying to
avoid. Combat will just get you killed!
Even then, the pattern that we've experienced in game is this:
* Below level 5, the PCs short rest after 1-2 combats. There are between 1 and 4 encounters per long rest.
* Beginning about level 5, the PCs short rest roughly once every two long rests. That is, once every other day. As the game progresses, the number of long rests between each short rest gets longer and longer. By the time we were level 15, I think we'd had 2 short rests since level 10. There are between 1 and 3 encounters per long rest.
Once the players get enough survivability, they
never short rest. So we've got 6-8 PCs and most encounters are hard, deadly, or higher. Inevitably, someone spends a lot of resources and wants to
long rest. That's why we have between 1 and 3 encounters per day. The only problem this causes is that short rest classes get shafted because they don't get to short rest. Basically, the rule is: Don't play Warlock, Fighter, or Monk and expect to get your powers back during the day. Either people in the party will consume enough long rest resources to prompt a long rest, or they don't and not enough PCs will want (or need) to stop. The short rest players just suck it up.
What stops the party from resting? Not much. Threat of ambush is really all there is, but
leomund's tiny hut and
rope trick circumvent much of that. Unless the PCs try to do something genuinely stupid like set up camp inside the enemy stronghold, they can find a pretty safe spot to steal a rest (long or short). We never kick in the door of a keep, kill everybody on the ground floor, and then ignore the rest of the levels while we take a short rest. That just never happens. If we need to rest that badly, we withdraw and regroup. Mainly it's just that short rests don't do enough to bother with. When you're level 8 with 14 Con, you've got about 60 hp max. With 10 out of 60hp you can freely recover only 4d8+8 (26) with Hit Dice over a short rest -- the rest of your Hit Dice don't come back for 2 days. [This has made me wonder multiple times if it would encourage more short rests if long rests just recovered all Hit Dice.] That gets you to 36 out of 60, which is still not good. Even then, half the party will often only have spent resources that are long rest refresh anyways.
This has led to my conclusion that short rests as a mechanic and as they exist in the game, don't work well. Maybe the designers intended which type of rest to be made by the party to be an "interesting choice," but they're really not. Most of the time, it's obvious that you need to long rest, and the other times, you just get the party disagreeing because one person needs a long rest, one needs just a short, and one is still undamaged with nothing lost. Adding time pressures doesn't really make for an interesting choice, either; it just makes not short resting the only choice. Including mechanics that encourage party disagreement is not interesting, and when you have one class with almost all their mechanics set to short rest, and another with all their almost mechanics set to long rest, the outcome should be pretty obvious. Short rests are not rewarding enough to all classes, or long rests aren't difficult enough to accomplish, for short rests to be truly worthwhile. [Again, this is in the context of significantly higher encounter difficulty.]
I know some people have switched to long rests requiring comfortable accommodations (shelter, warmth or fire, proper beds, good food, a bath, etc.) such that long resting basically requires a town or roadside inn to accomplish and everything you can do with a tent or in a dungeon is going to be a short rest. That does work, but it makes short rest classes a lot more compelling than long rest ones, and also might really limit your options. I can't imagine playing Out of the Abyss that way, for example. It's a lot more grim 'n gritty play style, though, and I am considering it for an open wilderness sandbox campaign.
I know other people have converted all short rest abilities to long rest by giving them 2x or 3x as many uses as a long rest ability. Others have gone the other way, and wanted to eliminate long rest and make all spellcasters built on the Warlock class model. (I think that would be a terrible disservice, because the Warlock class model is very broad but also very, very shallow. I also think the game would stop feeling like D&D.)