Wednesday, February 8, 2012

Critical Hits in D&D

This post over at Tenkar's Tavern (excellent and highly recommended, by the way) got me to thinking about critical hits in D&D, any flavor, including clones thereof. As a player, and as a DM who likes to see smiling player faces, I love critical hits. Nothing is quite so exciting as getting your ass handed to you and all of a sudden rolling that beautiful 20 (we always yelled "NATCH!"). You may still get your ass handed to you, but that crit insured that the other guy knew he had been in a fight.

Of course, we handled it differently over the years. There were hit location tables, confirming crits, secondary tables with more abstract things than direct hit locations (Bleeder!), and the ubiquitous double damage. It was the double damage we generally used most often. It was faster, less fiddly, and quite a visceral experience to slap some jackass with 28 points of damage.

The enemy could score crits on the players, too, and there was always the dreaded fumble. This usually was simply fall or drop your weapon. Either way, you basically missed your next turn.

I'm not such a fan of critical hits these days, and I'll tell you why.

  • D&D combat models results. It's roots are in a wargame, and a wargame is concerned with who wins, not how. Any sort of critical hit system intrudes on the model.
  • Combat is conducted with a d20. It is based on linear probability, which means the stable boy and the war hero will crit 5% of the time. Now, I know that confirming your crit mitigates this, but who wants to fiddle with another roll? And have you ever seen a player in a desperate situation that failed to confirm? It isn't pretty.
  • The roll to-hit is simply that, a roll to-hit. The only possible rationale for basing a critical hit from that would be to assume that a high roll (a 20) could indicate a hit to a more vital spot.
  • To me, it really should be the damage roll that indicates a crit. I have no system for this because I do not desire one. I am merely speaking hypothetically. Exploding damage dice or something. I just think that it is silly to roll a 20, double the damage, roll like a 2, and end up with a "critical hit" doing maybe 6 points. A damage roll that indicates max damage is more critical than that.

I am an old-schooler when it comes to weapon damage. I love the idea of all weapons doing d6. So, and this is completely off-the-cuff as I type this, I think if  was going to introduce crits, I would just allow an additional d6. Maybe. I don't know. I just know we are trying to model outcomes and introducing subsystems like critical hits into the model engine are doomed to either fail or disrupt the engine.


  1. Nothing ruins rpgs for me more than thinking about 'the maths' or 'the probabilities'.

    What has brought you more fun at the table: You or one of your friends rolling a 20 and yelling 'Natch' or realising that a 5 is as significant as 20 when using a d20 and the stable boy has as much chance to crit as the vet then removing crits and the chance to have some actual 'at the table' fun?

    Why not work up a fun table with a list of increasingly cool crits (or dire fumbles) then add/subtract the character's level + any to hit mods from stats or magic items to the result. That way you have a appreciably better effect for the vet than the stable boy, and more fun. "NATCH!"

    tl;dr: bah math, bah probabilities, bah mutter, mutter hey look I just rolled a fucking 20!

  2. This comment has been removed by the author.

  3. Well put, sir. Sometimes I do digress into theory, I'll admit. The actual fun at the table is something I haven't had in a couple of years, so I do tend to get mired in Theory and Philosophy, and forget the name of the game is Dungeons & Dragons. It's one of the many perils of not having a group. I don't have anybody to smack me on the back of the head and scream, "Just roll some fucking dice, you wanker!" Thanks for the "reality" check, Lee.

  4. Not so much a reality check for you as a personal bugbear for me. I just can't understand why anyone want's the dice to be more important in the game than 'what do I need to roll?' I want the fiction of the game at the forefront not the numbers.

    Anyway, have you looked at ConstantCon and gmail+ as a way to play, David? I'm playing in my first game tomorrow so I'll find if it works for me. If it does I'll prolly want to run some games.

  5. Actually, I'm in the camp that probabilities are important. It's not something the players need to worry about, but when you're designing a game, even just bolting on a critical hit rule, you need to think about how often what kind of result will crop up. Naive attempts to add crits to D&D will often end up killing the PCs out of all proportion to the added fun of whacking a goblin for 18 points...particularly when 6 would have killed it anyway.
    If you don't want to delve into the math and theory, it's simplest to make it crits are something only PCs can roll. To avoid the "Woohoo! Natch! Crap! I got a one" you either forget the crit on the to-hit and make it on the damage roll (e.g. with exploding dice), or make it something like crit = automatic max damage. The nice thing about both of those (getting back to the math) is they change the expected outcome very little, which means than on the whole you don't need to worry about it when it comes to eyeballing how tough an encounter is.

  6. Hi,

    Apologies for the off-topic comment, but I couldn't find a contact email for you.

    I've recently put out an ebook of my writing, called The New Death and others. It's mostly short stories, with some obvious gamer-interest material. For example I have a story inspired by OD&D elves, as well as poems which retell Robert E Howard's King Kull story The Mirrors of Tuzun Thune and HP Lovecraft's Under the Pyramids.

    I was wondering if you'd be interested in doing a review on your blog (either a normal book review, or a review of its suitability as gaming inspiration).

    If so, please let me know your email, and what file format is easiest for you, and I'll send you a free copy. You can email me ( or reply to this thread.

    You can download a sample from Smashwords:

    I'll also link to your review from my blog.


  7. My favorite method back in my 2E days was a choice for the player between double or full damage (both before bonuses). So the player gets to make a quick little risk/reward calculation: do they want the guaranteed full damage or do they want the the chance of a truly awesome (double damage hit) at the risk of potentially doing less damage than a normal hit. The probabilistic expected value is of course the same between the two methods. But this simple little choice always seemed to give players (and I'm including myself here) so much joy.

    To be honest, I have no memory if we used these rules for monsters also. I think we did, and the lethality of our games was not especially high.