Flat critical failure percentages

Thursday, 29. July 2010

One of the things that I’ve always hated in various d20 and d100 based systems is their method of critical failures.  No matter how skilled or unskilled a character is the chance of failure is always flat.  Be it 5% in a D20 system, or a varying 2-5% in a D100 system.

Most of us have tasks, jobs and skills that we are very proficient at either from hobbies or from work.  I happen to do a fair amount of technical work on equipment.  If I had anywhere near a 5% failure rate even under stress, people would be dead and I would be fired in a heartbeat. It’s probably well under 1% with actual failure rates perhaps slightly higher, but still underneath it. I’m sure everyone agrees a good amount of their work is similar, with a very low rate as well.

Now, very delicate work in hostile conditions? Probably higher depending on interruptions and the tools the character in question has to work with.  Higher still if they’re under attack or under a time limit (Bomb ticking away perhaps?) Character dosen’t care?  Higher, so on and so forth.

You can get this to a point by adjusting failure %’s flat out.  You are of course, limited to your die’s resolution… 5% increments for a d20, 1% for d100…  Even doing that however, it still leaves that annoying flat failure percentage.  No matter the skill of the character in question, the chance of critically screwing up remains the same.

The simplest solution that I can work with is to start using small dice pools.  I say small because I really don’t play games to roll dice.  If I did that I’d just play Yatzee or maybe craps.    For instance in the case of the d20.. have another die in there for a failure die. Your failure die can be tuned to whatever you feel the % needs to be.

Crit Failure on     Crit Failure
1 no fail die            5%
1, fail die 1-10       2.5%
1, fail die 1-5          1.25%

Another (more preferred option, in my case) is to use something similar to what heavy gear uses. Your skill levels determine the dice you throw, and the highest is kept. If all come up 1′s, crit failure. Standard dice used is d2, and throwing 2 of them is considered average. 3 is professional, etc. In addition to dropping the fail rate, it also consistently moves the average roll up making trivial and nontrivial tasks easier. I realize its still a set % get a certain number, but it still feels less “chancy” then relying on one die for some reason.

Mind you with the tables below that any skill above 3 is incredibly rare
Dice    Thrown Crit Fail %   

1          16.7%                            
2          2.77%                          
3          0.46%
4          0.07%
5          0.012%

The system tends to use compared rolls, so crit failures even in combat aren’t necessarily lethal.  In addition, the system allows you to burn small amount of experience to save your character in a bind, in case there is a critical failure.  All in all, very well laid out and it feels fairer for players involved to be cheated less by dice.

Thoughts on the matter?

Technorati Tags: , , ,

2 Responses to “Flat critical failure percentages”



  1. faustusnotes Says:

    I switched my d&d3.5 games to 2d10 without changing anything else, maybe 5 or 8 years ago. I think I made the switch after a particularly egregious series of 20s. 2d10 has the advantage of weighting the probability distribution in the middle, so that the “average” value actually corresponds with the most likely value (not the case with d20s). It also means that as the number you need to beat increases, your chances decrease rapidly. 19 or 20 on a d20 is a 10% chance of success; on 2d10 it’s 3%.

    I then further adjusted the rules to reduce criticals, by adopting the Rolemaster critical success rules, where you roll again and add, and only get a success if you beat the original target DC. So if your target DC is 28 and you roll 20, and you have a skill of 2, you still need to roll 6 or more on the second roll. This is like critical threats in d&d3.5, but no guarantee of success when you’re attempting something with a DC well beyond your skill.

    I considered switching to 3d8, so that the chance of a fumble is even lower, but I like fumble rates of 1%. Also, I use the take-10 rules in most non-combat situations, and don’t allow fumbles where there is no pressure (time or combat-related), except for hazardous activities and magic.

    For anything to do with skill systems, I cannot recommend strongly enough an initial education in Rolemaster skills. Skill resolution in RM is the most complex and detailed I’ve seen, and it works quite well. Once you’ve played around in RM, seen all its good and bad points (it has many of both) you can easily port the good points to a different system, and improve it immeasurably.

    For example, RM has partial failure rules. I introduced these in D&D3.5 by allowing someone to get a partial success if they roll below the DC but above (DC-level). So if your DC is 15 and you’re 2nd level, you get partial success on a 13 or 14. Partial success gives a lot of opportunities for role-playing, adding tension, etc. so I like to use it. Rolemaster may have been a beast of a game, but its fundamental ideas were well worth stealing.



  2. Grey Says:

    I’ll definitely have to take a look at the RM skill system then. My background is primarily D&D of all iterations, with some Palladium, WoD, HERO, Heavy Gear and 7th Sea thrown in. Graduated success and failure is definetely a good thing to work into any system, moreso when you’re working an averaging system such as you describe above.

Leave a Reply

Spam Protection by WP-SpamFree