On Criticals

23 posts / 0 new
Last post
This draws on some of the same logic as this thread on [thread=950773]armour and DR[/thread].

Currently, criticals are a bit tricky to resolve. Threats are easy, but I accept the 4e argument that the confirmation roll can be a downer. In contrast, the "roll a most (but not all) of your damage multiple times" is in practice surprisingly clunky to actually get right, even for the mathematically skilled.

The other issue with crits is that high-mult crits in combination with high damage can be swingy. One greataxe hit for 2d6+15 (not out of the question with power-attack) per round might not be too bad, but crit for 6d6+45 is nasty - that could be an insta-kill on a 4th level character. In contrast, critting with your 1d4+1 dagger means 2d4+2, which is still almost nothing.

I wonder if a better mechanic is for crit bonus damage to be "constant", at least for a given PC at a given level. For example, a crit could deal +10, or to scale better with level it could be 5 + 2x level, or some variation of the above. This has a few advantages:
* math to calculate extra damage on fly is easy (add one fixed value)
* less swingy
* even small weapons can deal good crits, which makes them special for everyone.
Here's a simple level-scaled mechanic: add your unmodified BAB to the damage. (No confirmation roll or anything.)

Here's another one: if using armor as DR, "crits" ignore armor.

-- Alex
Here's a simple level-scaled mechanic: add your unmodified BAB to the damage. (No confirmation roll or anything.)

That sorta works, but is IMO on the weak side (plus certain classes get no benefit from crits at level 1).

No confirmation roll (as proposed for 4e) does change the critical math a little. In 3.5, a 20/x2 critical increased mean damage by 5%, 19-20/x2 or 20/x3 by 10%, and so on.

An auto-crit on 20 changes the maths, since it adds a flat 5% chance of dealing crit damage, which (assuming double damage) could increase mean damage by up to 100% (need 20 to hit).
Here's another one: if using armor as DR, "crits" ignore armor.

Well, I think the "armour as DR" model is inherently poor for any significant DR. Having crits ignore this DR is an interesting mechanic, but suffers from scaling widely depending on whether the opponent is wearing armour (eg: "you can't crit monks!"). The modelling for this is also a bit wacky for lightly armoured targets: "I have to crit to hit an unarmoured location?".
I could see a crit as just adding your BAB to your damage. Or your Str mod. Or just a quick X2. Don't forget the blows that bypass AC. Hit even if they miss, and area effects. Now that everything gets a crit they're gonna have to simplify the mechanic.
I am greatly disturbed by the no confirmation roll. So when a kobold can only hit your heavily armored PC on a '20' he is in fact critting everytime. Maybe that balances out since the kobolds may only hit once a battle before dying, but its still feels wrong.
Need a laugh? Visit the one and only Greyhawk Comic!
I am greatly disturbed by the no confirmation roll. So when a kobold can only hit your heavily armored PC on a '20' he is in fact critting everytime. Maybe that balances out since the kobolds may only hit once a battle before dying, but its still feels wrong.

It's only a 5% chance. I'm cool with it.
I was actually thinking as I opened up this thread that reducing the critical multiplier could be a good move. Big crits are fun in a sense, but they're also swingy and life-ending for PCs. I was tempted to say that it'd be good to reduce it down to x1.5 instead of x2, but there's a lot to like about the +10 suggestion as well.

* Small weapons crit 'harder' - actually probably still hitting for less, but the damage increase is bigger for them on a percentage basis
* You can't crit for practically no damage. Ever crit with a hit for 1d4-1 in today's game? Yeah....
* +10 damage is always a good boost in today's damage numbers, and will be noticable
* +10 damage isn't so much that it will instantly gib PCs after the first few levels.
* If you really wanted to make some weapons do more critical damage than others, you can always give them a different critical damage modifier.
* Critical hit damage multiplications is another source of slowdown in combat.

Having a critical hit ignore DR as well sounds like a good move to remove those 'Oops! It still did no damage!' moments.

I'm not certain that I like the sound of 'no confirmation rolls' either, but I'd have to know more about the ruleset regarding criticals in 4th edition to actually make a judgment.
The +BAB to damage would be cool.
I thought i had seen somewhere that offensive spells will be critting too. Just adding the BAB doens't seem like an adequate solution if you have casters critting.
I thought i had seen somewhere that offensive spells will be critting too. Just adding the BAB doens't seem like an adequate solution if you have casters critting.

+Caster Level for spells?
I have to see the math but it seem to solve some of the problems with too deadly criticals at low levels.
I wonder if a better mechanic is for crit bonus damage to be "constant", at least for a given PC at a given level. For example, a crit could deal +10, or to scale better with level it could be 5 + 2x level, or some variation of the above. This has a few advantages:
* math to calculate extra damage on fly is easy (add one fixed value)
* less swingy
* even small weapons can deal good crits, which makes them special for everyone.

Good post. As I read down through everybody's comments I liked the idea of critical hit damage doing 'the constant' + BAB. I also think the weapon itself should have some effect on the amount. So what if that 'constant' was a number based on the weapon?

Since all weapon's crit on a 20 only, you could have a short sword's critical hit info read something like '+7' while a battle axe might be a '+11' (numbers just pulled out of thin air for example's sake). Then you would add your BAB (and maybe STR again?).

The Piazza A renaissance of the Old Worlds. Where any setting can be explored, any rules system discussed, and any combination of the two brought to life.

I thought i had seen somewhere that offensive spells will be critting too. Just adding the BAB doens't seem like an adequate solution if you have casters critting.

Though if casters are rolling to hit, they probably have a Spell Attack Bonus, so instead of +BAB, it's +SAB.
I honestly never had a problem with how crits worked before, I liked the confirmation, because as one poster stated, it kept the Kobold who could only hit on a 20 at all from critting every time, which doesn't make much sense to me. I also liked the variable multipliers, its really not that difficult of a math problem, and if it is, get smarter people to play with, but it wouldn't make sense to me that a greataxe wouldn't be critting for much more than a dagger. That's of course my opinion, but I had no issues with how crits worked out in 3.5.
I honestly never had a problem with how crits worked before

I can't say that I really have a problem with the way crits work in 3.5, rather, I think they could be done better. I do like the confirmation roll ideology but maybe that should effect the degree of the critical hit.

What I like about the concept we're exploring here is that a 'critical hit' should be critical regardless of the weapon. A critical from a dagger sould be just as critical to your survival as one from a large battle axe. Sure the axe is usually more life-threatening (i.e. damage-producing), once criticals are made the damage should be considered, well... critical, regardless of level.

The Piazza A renaissance of the Old Worlds. Where any setting can be explored, any rules system discussed, and any combination of the two brought to life.

I thought I recalled reading that 4E crits will natural 20 only and will be x2 damage, regardless of weapon type?

If so, it makes sense that once you do away with variable crit multipliers, you also need to do away with variable threat ranges, and vice versa.
It's an orthogonal discussion, Colmarr. 4e is also planning to do away with the confirmation roll. Although the confirmation roll is mathematically sound, there are legitimate reasons to remove it:

(1) It's disappointing to roll a 20 and then fail to crit.
(2) It's an extra mechanic which takes extra resolution time.

My issue is that if we're doing away with the confirmation roll, this means that you crit on 5% of your attacks rather than 5% of your hits. This changes the maths significantly, especially on targets with relatively high AC. Further, the actual mechanics of "roll double damage" are clumsy to figure out if they aren't already written down, especially since 3.5 only doubles some of your damage.

A clean way to work around all this is to make crits add constant damage (possibly scaled by level). By "constant", I mean "independent of the attack used to crit". I'd be happy with 10+level or the like, but one could do 2d6+level if there was this pressing need to roll more dice. *shrug* It has the nice effect that critting with a weak attack actually means something and critting with a strong attack isn't risking an insta-kill (ie makes combat less "swingy").

This isn't to say that you couldn't have a few special-crit weapons, that, say, "deal +5 damage on a crit". That's a nice, easy to implement, way of keeping the idea of special crit weapons while still making crits easier and faster.
It's an orthogonal discussion, Colmarr.

Oh, my bad :embarrass . It wasn't clear (to me at least) from the OP.

My issue is that if we're doing away with the confirmation roll, this means that you crit on 5% of your attacks rather than 5% of your hits.

I'm not entirely sure that I agree we were ever critting on 5% of our hits, but I won't argue the specific math. You were always better at it than me

We (as in "you and I") agree that there are valid game-play reasons for removing the confirmation roll. I don't necessarily think it has anything to do with player disappointment. I side more with the argument that "you hit! roll and see if you can hit again for more damage" is an ugly and time-consuming mechanic.

This changes the maths significantly, especially on targets with relatively high AC.

It certainly changes the maths, but I don't think anyone has done any hard calculations on the effect the new rule will have on damage dealt in an average combat. The lack of a confirmation roll will increase crit probability. "Natural 20 only" will for many popular 3.5 weapons decrease crit probability. And x2 crit damage will likewise lower damage dealt on a crit by many popular weapons.

I don't think there's anything inherently wrong with changing the math (and I'm not sure that you're arguing that it is wrong to change it).

The only truly offensive possibility on the released information is the "I need a 20, therefore every hit is a critical!" phenomenon. I trust Wizards to deal with that. It's as easy as inserting a sentence to the effect that you can't crit against an opponent if you need a 20 to hit.

Yes, needing a 19 and critting 50% of the time is egregious from a "realism" point of view. But if that is the "loophole" in a crit system that removes all of the book-keeping of multiple threat ranges, confirmation rolls, and damage multipliers, I'm all for it.

Further, the actual mechanics of "roll double damage" are clumsy to figure out if they aren't already written down, especially since 3.5 only doubles some of your damage.

Have they actually said this; that on a crit you will roll double damage. If so, I agree with you. It's an ugly mechanic. Do double damage is much easier to handle and more in line with the streamlining that D&D seems to be undergoing.

A clean way to work around all this is to make crits add constant damage (possibly scaled by level). By "constant", I mean "independent of the attack used to crit". I'd be happy with 10+level or the like, but one could do 2d6+level if there was this pressing need to roll more dice. *shrug* It has the nice effect that critting with a weak attack actually means something and critting with a strong attack isn't risking an insta-kill (ie makes combat less "swingy").

Agreed.
I am greatly disturbed by the no confirmation roll. So when a kobold can only hit your heavily armored PC on a '20' he is in fact critting everytime. Maybe that balances out since the kobolds may only hit once a battle before dying, but its still feels wrong.

Wasn't there a rule that you could only crit on a 20 if you had to roll 19 or less to hit? Like if you had to roll a 20 to hit, rolling a 20 would mean you just hit normally, and did not crit. So creatures that need to roll 20 to hit would never have the skill to crit you. I think I might have read this somewhere.
I don't like how criticals are basically this random chance of a big explosive hit. For tactical combat, I'd prefer something that scales more gradually, like with degrees of success of something. Unfortunately, d20 doesn't do degrees of success too well (you can certainly do it, but it's a lot of "how high is this die roll?" stuff that gets annoying very fast).

-- Alex
Scaling is a key issue here.

Natural 20 is crit. Does not scale. Crit chance is always 5%, no matter what. Mechanistically simple.

Natural 20 + confirmation roll. Crit chance does scale with attack bonus, but 5% is the maximum (thus scaling in a relatively small interval). Clumsy mechanic.

Degree of success (e.g. hit by more than 10 is a crit) This scales very well. High level fighters can do more crit than low level ones. BUT at low level there are would be very few criticals (maybe none), and at higher level too many of your attack will be a critical. This is unfun.

Now I like the scaleing damage idea. While the chance of critical might not scale, its effect can. So while having the same 5% chance of critting (1st method) but dealing extra damage based on BAB, high level fighters become more able than lower level ones. Also critical won't be so deadly at low levels, but would still do meaningfully more damage (even +1 is important when the enemies have 1 HD)
I like rolling for a crit confirmation. Yes failing is disapointing, but so is failing to hit. So lets make all attack rolls auto hit too then. PLEASE!

It does take more time, but one extra roll 5% of the time. I'm not seeing it slow down my table. When it does happen, it adds suspense, which isn't a bad thing.

What slows down the table more is counting up all the damage dice. Maybe adding your classes hit die to your damage. So a d10 for a fighter, d12 for barb, d4 for wiz.

I'll be sad to see crit ranges and multipliers removed from weapons. Hopefully they put something in 4.0 to make the weapons feel different. How is the heavy pick going to be different from the great sword now?
Wasn't there a rule that you could only crit on a 20 if you had to roll 19 or less to hit? Like if you had to roll a 20 to hit, rolling a 20 would mean you just hit normally, and did not crit. So creatures that need to roll 20 to hit would never have the skill to crit you. I think I might have read this somewhere.

If they are getting rid of crit confirmation roles then this would be a very sound rule, I used to like the three 20s=instant death rule, but I think it's a bit daft now.
Sign In to post comments