Death and Dying

Character death is one of the ultimate threats in any RPG, and D&D is no exception. Besides the obvious, um, “inconveniences” that death might cause your character and his allies in both the short and long term—inconveniences which vary based on your level, the current situation, and of course your attachment to that particular character—death is a mark of failure. In some hard-to-explain but very real way, a dead character symbolizes that you just “lost” at D&D. That can prove a bitter pill for many players, and in my experience is even more frustrating than paying for a resurrection.

What We Hated
Early in the design process, Rob, James, and I identified a number of ways that we were unsatisified with D&D’s current death and dying rules. For example, we strongly disliked the inability of 3rd Edition D&D’s negative-hit-point model to deal with combat at higher levels—once the monsters are reliably dealing 15 or 20 points of damage with each attack, the chance of a character going straight from “alive and kicking” to “time to go through his pockets for loose change” was exceedingly high; effectively, the -1 to -9 “dying” range was meaningless. Ask any high-level fighter whether he’d prefer the second-to-last attack from a monster to leave him at 1 hp or -1 hp; I’d put odds on unconsciousness, and how lame is that?

Among other problems, this also meant that characters effectively had no way to “lose” a combat except by being killed. This removes a lot of dramatic possibilities for the story—for instance, the classic scene of the characters being captured and thrown in a cell from which they have to escape using only their wits and a pack of chewing gum (or whatever).

On top of all that, the game added a complex state of being at exactly 0 hp, which wasn’t quite like being fully capable but also wasn’t quite dying. Honestly, though, how often does any character actually get reduced to exactly 0 hp? Why did the game need a condition that existed at exactly one spot on the big, broad range of hit point possibilities?

What We Wanted
We wanted a death and dying system that added fun and tension at the table, scaled well to any level of play, and created the threat of PC mortality (without delivering on that threat as often as 3rd Edition did).

Characters had to feel that death was a possibility in order for combat to feel meaningful. If it seems impossible to be killed, much of the tension of combat disappears. However, if the majority of combats result in death (as is the case for a lot of high-level play in previous editions), the game is forced to reclassify death as a trivial obstacle in order to remain playable. 3rd Edition accomplished this with popular spells such as close wounds, delay death, and revivify—mandatory staples of any high-level cleric’s arsenal due purely to the commonality of death. But that removes the tension, and now what’s the point of death at all?

The system also had to be simple to remember and adjudicate at the table. Being able to keep the rule in your head is important, because you don’t want to be bogging the game down flipping through a book when a character is clinging to life by a thread—that should be high-tension time, not slowdown time!

Finally, it had to be believable within the heroic-fantasy milieu of D&D. (Believability isn’t the same thing as realism—an error which has ruined more games than I can count.) Put another way, it had to feel like D&D—one of those tricky “you know it when you see it” things.

What We Did About It
Back in 2005, this was obviously a much lower priority than, say, creating the new model for how classes and races worked, so we put it on the back burner to simmer. As the months passed, we and other designers proposed various models that tried to solve the conundrums set out above, varying from exceedingly abstract to witheringly simulationist. We playtested every model, from death tracks to life points, each time learning something different about what worked or didn’t work. A few times, we even temporarily settled on a solution, claiming that the playtesters only needed time to get used to our radical new ideas.

Thankfully, our awakening came well before we released the game (or even before widescale playtesting began, for that matter). Despite some quite elegant concepts, none of our radical new ideas met all the criteria necessary, including simplicity, playability, fun, and believability.

The system had to be at least as simple to remember and at least as easy to play as what already existed. For all their other flaws, negative hit points are pretty easy to use, and they work well with the existing hit-point system.

It had to be at least as much fun as what already existed, and it had to be at least as believable as what already existed. In ideal situations, negative hit points create fun tension at the table, and they’re reasonably believable, at least within the heroic fantasy milieu of D&D, where characters are supposed to get the stuffing beaten out of them on a regular basis without serious consequences.

Every one of our new ideas failed to meet at least one of those criteria. Maybe they were playable but too abstract to feel fun or believable, or they were believable but too complicated to remember. Nothing worked, and I admit we experienced a couple of freak-out moments behind closed doors.

The Breakthrough

Eventually we got it through our heads that there wasn’t a radical new game mechanic just waiting to be discovered that would revolutionize the narrow window between life and death in D&D. What we really needed to do was just widen the window, reframe it, and maybe put in an extra pane for insulation. (OK, that analogy went off the tracks, but its heart was in the right place.)

Characters still use a negative hit point threshold to determine when they move from “unconscious and dying” to “all-the-way-dead,” but now that threshold scales with their level (or more specifically, with their hit point total). A character with 30 hit points (such as a low-level cleric) dies when he reaches -15 hit points, while the 15th-level fighter with 120 hp isn’t killed until he’s reduced to -60 hit points.

That may seem like an unreachable number, but it’s important to remember that monsters, like characters, aren’t piling on as many attacks on their turn as in 3rd Edition. At 15th level, that fighter might face a tough brute capable of dishing out 25 or 30 points of damage with its best attack… or nearly twice that on a crit. The threat of “alive-to-negative-everything” on a single hit remains in play, but it’s much less common than in the previous edition. That puts that bit of tension back where it belongs.

The new system also retains the “unconscious character bleeding out” concept, but for obvious reasons speeds it along a bit. (There’s not really any tension watching that 15th-level fighter bleed out at a rate of 1 hp per round for 30 or 40 rounds.) Thanks to some clever abstractions, the new system also removes the predictability of the current death timer. (“OK, Regdar’s at -2 hp, so we have 8 rounds to get to him. Yawn… time for a nap.”)

It’s also less costly to bring dying characters back into the fight now—there’s no “negative hit point tax” that you have to pay out of the healing delivered by your cure serious wounds prayer. That helps ensure that a character who was healed from unconsciousness isn’t in an immediate threat of going right back there (and you’ll never again have the “I fed Jozan a potion of healing but he’s still at negative hit points” disappointment).

Monsters don’t need or use this system unless the DM has special reason to do so. A monster at 0 hp is dead, and you don’t have to worry about wandering around the battlefield stabbing all your unconscious foes. (I’m sure my table isn’t the only place that happens.) We’ve talked elsewhere about some of the bogus parallelism that can lead to bad game design—such as all monsters having to follow character creation rules, even though they’re supposed to be foes to kill, not player characters—this is just another example of the game escaping that trap. Sure, a DM can decide for dramatic reasons that a notable NPC or monster might linger on after being defeated. Maybe a dying enemy survives to deliver a final warning or curse before expiring, or at the end of a fight the PCs discover a bloody trail leading away from where the evil warlock fell, but those will be significant, story-based exceptions to the norm.

Oh, and speaking of zero hit points? You’re unconscious and dying, just like every new player expects it should be. It’s not as harsh as the “dead at 0 hp” rule of the original D&D game, but it’s still not a place you want to be for long!

Try It Now!
If you want to try out a version of this system in your current game, try the following house rule. It’s not quite the 4th Edition system, but it should give you an idea of how it’ll feel.

1) At 0 hp or less, you fall unconscious and are dying.
Any damage dealt to a dying character is applied normally, and might kill him if it reduces his hit points far enough (see #2).

2) Characters die when their negative hit point total reaches -10 or one-quarter of their full normal hit points, whichever is a larger value.
This is less than a 4th Edition character would have, but each monster attack is dealing a smaller fraction of the character’s total hit points, so it should be reasonable. If it feels too small, increase it to one-third full normal hit points and try again.

3) If you’re dying at the end of your turn, roll 1d20.
Lower than 10: You get worse. If you get this result three times before you are healed or stabilized (as per the Heal skill), you die.
10-19: No change.
20: You get better! You wake up with hit points equal to one-quarter your full normal hit points.

4) If a character with negative hit points receives healing, he returns to 0 hp before any healing is applied.
In other words, he’ll wake up again with hit points equal to the healing provided by the effect—a cure light wounds spell for 7 hp will bring any dying character back to 7 hp, no matter what his negative hit point total had reached.)

5) A dying character who’s been stabilized (via the Heal skill) doesn’t roll a d20 at the end of his turn unless he takes more damage.