Critical Hits - why, and why not?

[Celebrim]

I started critical hits with a group that used the Dragon article.

It quickly occurred to me that critical hits favored NPC's over PC's. There is an unending stream of NPC's and it really doesn't matter how they die, but your PC can only get lucky for so long.

On the other hand, despite the fact that it is against their own interests, players themselves love criticals. Emotion trumps logic here. While arguably the game would be better without them, virtually every modern RPG embraces criticals because players enjoy them in the same way gamblers love to gamble.

Most RPGs that use criticals tend to have means for mitigating against them. For example, they tend to make criticals from NPC's less frequent and less powerful than those available to PC's - particularly compared to PC's that invest resources in generating criticals. In all D20 games, this is generally true in that most monsters can only generate 20/x2 criticals, where frequently you see players with 19-20/x3 or 17-20/x2 critical hits (or larger). 4e preserved this aspect to a large extent, and if anything over nerfed NPC criticals relative to player criticals resulting in broken math at higher levels were it was hard for a monster to challenge a PC party and things could get grindy if you weren't careful.

In my game, I further mitigated against criticals by the use of destiny points which could among other things cancel a critical hit made against a PC. This helps mitigate against a PC getting one shotted by a lucky orc with a great axe or any similar sort of situation (power attacking frost giant with a great axe, critical hit with a disintegrate spell, etc.).

Critical hits that cripple characters should be generally avoided. Maiming a PC is generally worse than killing one from an emotional standpoint. A maimed PC generally has to be abandoned unless magic to overcome the disability is so readily available that the maiming itself it meaningless anyway. Otherwise, a maimed character is effectively dead to the player but without catharsis. Losing a character always hurts, but if the character is dead the player has at least (hopefully) a good death story to remember and the excitement of character gen. Maimed characters that are retired haunt the process.

Maiming however realistic also makes combat so unpredictable as to make a joke of attempts to balance encounters. It's more or less impetus to avoid combat as a focus of play and the story, not the least of which is because playing out combat will probably also be burdensome as you try to resolve damage and its effects. And in general, it tends to make combat less fun, since the first telling blow tends to win the fight and reducing the fight to a death spiral of greater and greater gimpiness. So unless your goal is simulating a real era, I'd generally avoid critical hits (or just hits) that leave lasting non-ablative wounds. One alternative here however would be maiming as death mitigation, where the system tends to replace deaths with maiming specifically to allow for recovery in a system were raising the dead isn't usually possible but recovery from catastrophic injury generally is. I can see doing this in many science fiction settings, and in certain fantasy settings. In my own system, this maiming replaces death shows up in a few limited ways, notably replacing D20's standard 'massive damage' rule.

I generally agree with you that the 5e approach is logical. But as I said, the player preference here is an emotional one and not a logical one. It remains to be seen whether players will enjoy the less impactful critical or if we'll see house rules for more potent critical hits become more common.

 

[Umbran]

Gygax was writing a tight, gritty dungeon crawler. When you are writing a tense game the added swinginess caused by critical hits is a bad thing because it forces PCs to use much wider safety margins, and they have the dice rather than their foolhardiness to blame.

Except, you know, for those places were he wan't. Tomb of Horrors being a fine example of where swinginess is the source of tension, rather than a breaker of tension.

oD&D and 1e rightfully do not have critical hits.

Like Dragonlance, 2e should have had crits. It actively advocates fudging dice

Gygax explicitly speaks for fudging in 1e as well, so that's not a good measure for your analysis.

 

[N'raac]

Critical hits that cripple characters should be generally avoided. Maiming a PC is generally worse than killing one from an emotional standpoint. A maimed PC generally has to be abandoned unless magic to overcome the disability is so readily available that the maiming itself it meaningless anyway. Otherwise, a maimed character is effectively dead to the player but without catharsis. Losing a character always hurts, but if the character is dead the player has at least (hopefully) a good death story to remember and the excitement of character gen. Maimed characters that are retired haunt the process.

Maiming however realistic also makes combat so unpredictable as to make a joke of attempts to balance encounters. It's more or less impetus to avoid combat as a focus of play and the story, not the least of which is because playing out combat will probably also be burdensome as you try to resolve damage and its effects. And in general, it tends to make combat less fun, since the first telling blow tends to win the fight and reducing the fight to a death spiral of greater and greater gimpiness. So unless your goal is simulating a real era, I'd generally avoid critical hits (or just hits) that leave lasting non-ablative wounds.

I always found it odd that critical hit systems would sever limbs (or crush them into uselessness), but never break a bone, or just leave a long-lasting impairment to the limb like a deep slash or bruise. Their proponents would then tout the "increased realism" of the system.

 

[Neonchameleon]

Except, you know, for those places were he wan't. Tomb of Horrors being a fine example of where swinginess is the source of tension, rather than a breaker of tension.

Tomb of Horrors is generally a terrible example of anything. And is here too.

Firstly, it isn't so much swingy as binary. Alive/Dead with nothing in between. But more importantly this binary state is a response to player mistakes rather than to dice rolls.

Secondly and more importantly, it was created for two players (Rob Kuntz and Ernie Gygax) who were claiming that everything in Greyhawk was too easy. So Gygax made it as hard as he possibly could while still being "fair". Unfortunately for Tomb of Horrors, if you want to beat the dungeon, beat the designer. And both Rob Kuntz and Ernie Gygax had beaten Gary Gygax by then (which was why they found Greyhawk too easy) and ToH is an extreme case of "If you want to beat the dungeon, beat the designer" with a simple algorithm to get you through safely right up to Acecerak. Therefore not a single PC died in the first run through of Tomb of Horrors (although they teleported away from Acecerak himself after grabbing the loot on the grounds fighting was for chumps).

(Source: Mike Mornard on RPG.net)

It's raw player skill with harsh consequences for falling off the tightrope. Swinginess of dice have nothing to do with what it was designed for.

Gygax explicitly speaks for fudging in 1e as well, so that's not a good measure for your analysis.

Point.

 

[Celebrim]

I always found it odd that critical hit systems would sever limbs (or crush them into uselessness), but never break a bone, or just leave a long-lasting impairment to the limb like a deep slash or bruise. Their proponents would then tout the "increased realism" of the system.

No system that doesn't track blood lose, shock, and infection in detail has any claim on being 'realistic'.

What hit location/wound systems do actually do is something I don't have a good name for but usually call 'cinematic' out of long habit. Unfortunately that term has over the years come to mean something completely different than what I originally started using the term to mean decades ago. When most people use the term 'cinematic' they tend to mean, 'verisimilitude to the tropes of Hollywood action movie', and I think that's become more or less it's 'official' definition in PnP RPG's.

But by 'cinematic', what I mean is that the resolution stage of the 'proposition->fortune->resolution' cycle tends to suggest concrete images for the players to mentally imagine. So for example, one pretty good objection to hit points is that they are decidedly not cinematic. The resolution of the proposition, "I swing my sword", tends to be something like, "You do 6 damage to the orc." But this doesn't suggest any sort of concrete image at all, and in no way helps the players imagine the action they are involved in.

When the proponents of hit location/wound track systems praised them as realistic, what they were really saying in my opinion is that those systems were more "cinematic" (my definition) in that concrete propositions tended to create easily imagined results. Those results didn't have to be in any way 'realistic', they just had to be something that you could imagine happening without a lot of fiat invention.

If you roll back RPG discussions to the '80's, we didn't have a lot of vocabulary for discussing the design of an RPG and its effects on play. So what you tended to observe was people very loosely using 'realism' to mean a lot of vague but often very different sorts of things. This tended to lead to the fetishizing of 'realism' as the be all end all of design, as if every problem observed at the table or with the design could be cured by greater 'realism'. I don't think it was until people really hard core started playing 'realistic' systems, that they began to get an understanding that greater 'realism' was not only not necessarily the solution, but not even strictly speaking what they had wanted in the first place. I know that was the case for me.

 

[gamerprinter]

I always found it odd that critical hit systems would sever limbs (or crush them into uselessness), but never break a bone, or just leave a long-lasting impairment to the limb like a deep slash or bruise. Their proponents would then tout the "increased realism" of the system.

As an aside, I helped create the Bone-Breaker racial paragon class for kappa, for the Kaidan setting of Japanese horror (PFRPG) because it is part of kappa folklore in their skills at breaking bones of their opponents.

Basically the kappa gets a special single attack doing shattering strike which based on level of the kappa simulates broken bones in the reduction of movement (no lower than 5 movement), or at 5th level causes strength damage (simulating the breaking of an arm), at 10th level constitution damage (simulating broken ribs), or at 15th level as intelligence damage (simulating broken skull). The kappa can choose a lesser form of damage with any strike.

It doesn't perfectly simulate broken bones, but its isn't bad.

 

[The Crimson Binome]

I despise critical hits. It is one of the few issues on which I agree with Mr. Gygax.

They are entirely too random, and complicate the game unnecessarily. They increase the likelihood of PC death, but do very little to help the PCs. They make it much more difficult to weigh your options in combat, because there's always a small chance of a catastrophic failure.

It's just a bad rule, flat out.

 

[Raith5]

Crits are absolutely crucial to D&D IMO for the sake of theatre if nothing else. I think max damage for crits (as per 4e) is great, but the plus d 6 per plus of the weapon was very spiky and complicated things. It always felt like a strange mathematical tack on to increase the damage of PCs.

I played with critical fails in 3e with natural ones being a dex check to drop the weapon. I really didnt add much except that all the martial types had *numerous* back up weapons, because at high levels with all the attacks going on, it happened regularly. It was a silly house rule in hindsight.

 

[Celebrim]

I played with critical fails in 3e with natural ones being a dex check to drop the weapon. I really didnt add much except that all the martial types had *numerous* back up weapons, because at high levels with all the attacks going on, it happened regularly. It was a silly house rule in hindsight.

Fumbles are something you have to be very careful with, and you really shouldn't add them at all unless you fix spellcaster/martial balance better and are at least occasionally having fumbles for spells.

As you noticed, particularly because of iterative attacks, your fumble rules have to be constructed in such a way that higher level characters have less chance of fumbling per attack than lower level characters. That means that you need to make the DC's to avoid the fumble static but base the bonus to the roll on something that scales or likely scales. Basing it on an ability score like DEX, that for the most part doesn't scale, ensures the result that you observed. But if you base it on BAB, or Fort Saves, or suitable skill checks, then you can have very low chances of fumbling per attack.

The best reason for having fumbles is to increase the 'theatricality' (picking up on your term) of the combat. D&D combat is inherently low theatricality, in that hit points don't have a natural theatrical component. You don't want to have them often, but if they show up you want them to help create some of that missing theatricality. For that reason, unlike criticals, I think fumbles have to be diverse. If all a fumble represents is dropping your weapon, then the second or third time it happens it becomes redundant and ceases to add any drama to the fight.

My present fumble rules are working pretty well for making fumbles rare. If anything, I'd say they are working too well. Probably the strongest complaint is fumbles so rarely have an impact that they are almost not worth tracking. They are working slightly less well for adding drama. I've got too many fumble results like 'Stumble. You have a -4 circumstance penalty to AC until the start of your next turn.', that have too much of a purely numeric component and not enough of a visual component. These sorts of fumbles rarely have an effect on combat (in this case, the fumble is meaningless unless a monster attacks you in the next round) and don't really do a lot of work to create a scene. Overall I'm happy with my fumble rules, but I figure after a few more years of play testing I'll probably be able to make some tiny tweaks based off of the lessons learned. When fumbles have happened, assuming that they happen very rarely, they've often been fun for everyone so they are worth keeping.

One thing I think I will be adding is fumble chances for more different types of combat actions than just attacking. In particular, I'm looking hard at the 'Draw a weapon/take something from my pack' types of actions and thinking that a check of some sort to avoid fumbling that is warranted.

 

[Umbran]

Swinginess of dice have nothing to do with what it was designed for.

The *source* of the swinginess is irrelevant, in terms of the effect on "tension". In this context, tension is developed as resources are slowly whittled away, and the threat of reaching that last hit point increases.

Critical hits break this tension, in that they jump over the point the player *expects* the character to die. Anything that jumps past the resource-management aspect of the game will do that, dice or otherwise.

I am noting that doing so regularly replaces the resource-dwindling tension with another tension - a general fear of unexpected death. The presence of Save-or-Die effects in the early versions of the game do this already, in fact, without reference to particular modules. I think this kind of sinks the whole "Gary was building for this one form of tension" thing.