You know that DM that likes their game to be one which hero death matters? Who takes from their hat add-ons to resurrection and raise dead spells that effectively make them once-in-a-campaign events..?
Yes, I've even gone there. I've also consistently seen it backfire. You make a huge deal out of once-in-a-campaign-level resurrection, and, then that PC just up and dies again out of the blue, or another PC as or more deserving/central-to-the-plot/whatever does, or the player moves away or otherwise becomes unavailable.
Who does not shy away from employing save-or-die (or better, choose-wisely-or-die) scenarios, as long as they are well telegraphed to the players so to avoid gotchas.
That can tend to get you a game of Paranoid Fantasy Roleplaying rather than Heroic Fantasy Roleplaying - "not that there's anything wrong with that..." ;P
Something that has been happening for a while (it is not a 5e specific issue)... I have players that simply refuse, at any cost, to spend one turn or a higher level resource to rescue a comrade in bad conditions during a melee.
That actually is something that kinda came back with 5e, because healing resources, though perfectly adequate overall, are a little weak and combats are tuned to be fast, so staying on offense rather than standing up an ally (especially doing so 'inefficiently' with the low-hps restored by Healing Word) often seems the way to go. It's similar to the situation in 3.5, really. Out-of-combat healing is resource-efficient (cheap wands in 3.5, HD not useful for anything else in 5e) and using your action offensively is very effective (because of OP spells & combos in 3.x, and fast-combat-tuning in 5e).
I find it more annoying that by entitling themselves to be complete arses, they even manage to complicate some encounters, by not preemptively acting to protect one another, followed by a dropped ally, who subsequently loses their action and is one less threat for the enemies.
That sounds like a problem, but how do you "preemptively act to protect one another" in D&D? Especially in 5e, how is just focusing fire to reduce the threat posed by enemies by simply dropping them ASAP not the most obvious way to do that?
Using an action now to prevent an ally from dropping is perceived as a weaker strategy as acting before the fact to prevent the drop and the ally's losing an action
Not sure I see the distinction. If you mean healing an ally to prevent them from dropping, yes, it's seen as a weaker strategy both because healing can't generally keep up with damage inflicted and because of the next point...
Healing a fallen ally is cheaper, as any leftover damage does not get accounted for.
Definitely true in an analytic sense. It's more efficient to let the enemy 'waste' damage in overkilling an ally, then negating all that excess damage
When the fight is too close to the end, it seem better just to try for a finishing blow and them rescue any fallen comrades.
Certainly true. In 3.5 you could count on ending a fight very quickly and in 4e dying rules were pretty forgiving so you could count on an ally staying alive for a round or few while you did so.
In 5e, both are true, and "fight close to the end" could be the case quite early in the battle...
While I generally agree with the third bullet myself (even though I find it somewhat reckless in some circumstances), the first bullet really annoys me. It has over and again being detrimental to the party tactics, but selfishness usually takes over and they never learn the lesson, even when adventurers start dying as a result of a battle becoming tougher.
The second bullet is a result of the specific way health is abstracted in 5e, and I find it even more annoying, as it effectively introduces the "bouncing heroes" effect.
"Whack-a-mole healing," we call it. It's always been an issue in any past edition or variant that had any dying options beyond instant-death-at-0-hps, because healing an ally in combat could always be 'wasted' if that ally didn't take any more damage that fight. Heal-from-0 and very generous dying rules (being dropped is even less life-threatening in 5e than it was in 4e), certainly build upon that incentive to a greater degree than ever, though.
But it'd be very hard to get away from by making combat 'deadlier' or healing/raising rarer or weaker or harder to use, because that only makes those less attractive options compared to all-out 'nova' offense that leverages 5e's fast-combat design to drop enemies faster than they can drop allies, rather than keeping allies up.
To address these issues (mostly the first and second bullets), I could go full "evil" and just start using actions from the baddies to terminate fallen heroes as soon as these baddies notice the healing capabilities of the party. But, on the other hand, sometimes I feel like I am even more of an arse than my usual self by doing this. This "solution" has, though a subjective advantage of not needing any rules tweak.
A good case in point. By raising the stakes, you make the most expedient option (full out offense from the surprise round on) that much more attractive.
Then I thought about a second option. What if every time someone gets dropped, they will take at least some minutes to be able to recover consciousness?
Then once someone is dropped they can forget about being healed, because it's a 'wasted' action in the action economy (and a wasted slot), and the efficient way to go is to let the ally recover using HD after the fight. Trying to heal to stay ahead of enemy damage is also still a losing proposition, so, again, full-on offense would seem like the winning strategy, leading to rollovers that don't feel 'dangerous,' leading to the DM upping the difficulty...
I know it could potentially let Jimmy very bored as his selfless chevalier PC got down in the beginning of the combat trying to protect that frail spellcaster ally, but I think most of the time the party would go to extra pains to avoid this from happening as it would be very detrimental for the whole group.
What extra pains? Probably not pro-active healing while the tough PC is still up. Trying to take turns evenly-absorbing damage is theoretically efficient but tough to pull off and probably out of character for a lot of PCs.
Do anybody have another option that could address these points?
If you want to get away from whack-a-mole healing, you do, indeed, need to get rid of heal-from-0 as you suggest, track negatives, die at negative CON or 1/2 hps or something, and require healing to heal those negatives. But, you'd also need to make healing more potent across the board - for instance, when you're healed by magic, you can also spend 1 HD/spell level to get back more hps, and/or healing spells could use the target's HD size instead of their usual dice, and/or the caster's stat or the target's CON could add to hps recovered, etc... That way keeping an ally up with in-combat healing seems more viable vs healing only dropped allies or only stabilizing dropped allies and maximizing slots & actions available for offense.
Speaking of offense the other thing you might want to do is re-tune monster stats and combat rules to produce longer, more elaborate combats that can't be rolled over so quickly. It'd mean going against the 'fast combat' goal of 5e, which is among it's best-supported goals, but it would also weaken the all-offense 'nova' strategy.
Maybe I am just facing an odd problem. Some of my players have been complaining about that "bouncing heroes" feature for quite some time already
Not an unusual phenomenon. You can find several 'whack-a-mole' threads about it, here on ENWorld.
it really annoys me, as a DM, to see one of my players' PC die because the others are "gaming the game" too much (for lack of a better expression), or being just plain egotistical. Also, if I can somehow tweak the rules to change this dynamic, I will.
Nod. It's just that that tweak might, intuitively, be towards making the game less risky (it already seems pretty /easy/ but that's potentially a different quality than risky or deadly) rather than more.
Her rationale was that if everybody could barely survive, they could inexpensively recover HPs on a short rest, and her lay on hands was too valuable to be spent on anybody else, as she was the paladin, who could rescue everybody else if things went wrong.
She may not have been that wrong, if no one else could heal.
On yet another campaign, a light cleric would always try to concentrate all his spells on offense. This would even mean he would rather burn a 1st level slot at an enemy to deal 4d6 @ 70% hit chance (average damage expected ~10) in place of healing the rogue for 1d8+4 (average 8.5). Some might say this is sound as he could potentially drop the enemy, and the heal was smaller than the expected damage, but he would do this even when he didn't really have any expectation to drop an enemy (for instance when fighting a giant, while only halfway through the battle), while he knew by the rogue's remaining HPs that a 6-8 HP heal was good enough to keep the she standing for at least one more round
6-8 hps making a difference vs the damage dished out by a giant?
One observation on regard to healing amounts and damage amounts, when a party has a combined HP total of, let's say, 150, and the enemies have a combined HP total of 450 (not at all unusual at my table), to be able to win, the party must be able to output more than 3x as much damage as they can withstand.
Assuming no healing...
That would also mean that a healing of around 1/3 of the expected attack output usually breaks even in the end of the fight
That depends on the damage output of the enemy, too. If the healing buys an ally an extra round of attacking, and the ally is about twice as good at attacking as you would have been attacking with the same-level slot, maybe. If the heal stands a decent chance of making no difference (you don't know which wounded ally is going to be attacked next, heal one, and the other gets dropped, for instance, or you heal an ally and the next hit drops him, anyway).
Revivify is one exception to the resurrection spells that I just "charge" the expected from the book... Curiously, this one spell usually was kept on the paladin arsenal, and I think she felt really responsible to have this available for the sake of the others, if need would arise. ...It was just not very fun for the other guy that was holding the frontline with her when he saw that when she was getting the pounding, she would go full-in healing herself for some 70 HPs after she lost half her total, while in other fights he was taking the pounding and she would patiently wait for him to drop and only then she would heal him back some crumbs, so that he would proceed to drop and raise again and again. I does feel like good teamwork is what is missing...
It may seem to generate weird fiction, but what you describe is a pretty reasonable way of managing hp resources, given that the Paladin is the prime/only source of in-combat healing. She can't afford to drop, the most efficient use of her healing on behalf of allies is to stand them up when they've been overkilled to leverage the heal-from-0 rule while preserving their actions (though in some cases, that might not work out so well, depending on the initiative cycle).
When I run the game, I try not to make it too much of a game of guessing, as I find it a more rewarding experience when players make their choices well informed.
"Early healing" is kinda a guessing game. Who will be attacked next? Will they be hit? For how much damage?