Wednesday, December 4, 2013

Being even more elegant

The DnD Next blog post this week is the second installment on game design elegance.  One of the key points in it is a discussion about making rules local rather than global.  The idea of the article is that you should bury rules in individual classes, weapons, or other small categories rather than making them standard so that only the people who really need to know them have to learn them.  For example, they talk about how spellcasters have to concentrate to maintain some spells and being hit ends that concentration.  The idea is that only the spellcasters will need to know this and other people will not.

I think this is a very flawed approach to solving an old problem.  The problem is the extreme complexity of the 3rd edition combat rules, in particular the rules about attacks of opportunity.  Firing a ranged weapon provoked an attack of opportunity, drawing a weapon did not, but sheathing one did.  Standing up provoked an attack, 5 foot stepping did not, and you can make a special kind of check to avoid provoking when spellcasting.  This is only the tip of the iceberg and anyone who wanted to not get destroyed in combat needed to know these things.  However, the problem wasn't that the rules were in the Combat section but rather that they were flat out too complicated!  The basic rules of how combat works need to be super simple but people need to know what they are.

Continuing the example from above it is clear that absolutely everybody in combat needs to know that concentration spells are ruined when you get hit.  When the enemy wizard tries to turn your friend into stone you *need* to know that if you go over there and clobber the wizard the spell will fail.  It simply isn't enough for the wizard to know that they have to avoid getting hit; the brawlers have to know the rule too.  Another example is that ranged weapons are likely to have a property that makes them bad when used within an enemy's reach.  The proposed solution is to have that rule sitting in the section on that weapon itself instead of in a central location.  Again though everybody needs to know that running up to an archer and getting in their face is a good thing to do mechanically - it isn't reasonable for that knowledge to be only available to the archer!

The fact is that if a rule is useful and if a rule is necessary then it needs to be located in a place where everybody can read it.  If that means that there are too many rules then trim down or simplify the rules.  Hiding the rules away so that they are hard to find or so only a few people know what they are just means that some players are going to play terribly because they don't know how the game works and that isn't helping anyone.  Players who end up on the ass end of a rule they didn't know existed will not thank you for making that rule local rather than global.


  1. Obviously, these people aren't programmers. If they were, they'd know that what they're describing is considered not just inelegant, but unmaintainable.

  2. Your description of attacks of opportunity makes it sound very intuitive. Was that 3rd or 4th edition?

  3. My description was a very small sample of the 3rd edition rules, which taken in their entirety were far too complex for such a staple part of combat. 4th edition attacks of opportunity were quite reasonable and straightforward.

  4. bwross: I was thinking of programming too, and I think they actually modelled their rules after coding practices without understanding why those practices are good for coding and not good for user manuals. If I were coding a bow I might put how the bow works in the bow object, which works great when I have to reference the bow in other parts of the code, but not so great when a human being has to go and look at a particular page in a particular book every time they want to use a bow.