Follow That NPC!

One thing the D&D core rules don't cover is the chase scene. Various third-party supplements have tried to fill that particular hole, but I've never been satisfied with them. So -- it's brainstorming time!

First, a description of the status quo: if one character (the "predator") is chasing another (the "prey"), the character with the higher speed wins. In other words, an unencumbered human (speed 30 ft.) chasing an unencumbered halfling (speed 20 ft.) will always catch him; if the human were running away from the halfling, he'd always get away.

Why is this a problem? Mainly because it's boring and frustrating. It's all about absolutes; there's no tension, no possibility of surprise. It's flat-out unsatisfying.

How do we add the possibility of surprise? By adding randomness. By adding dice.

So, we turn it into, at minimum, an opposed check. But before we get ahead of ourselves, let's think: what kinds of chases are there?

1. The trivial chase. The predator is chasing his prey across an open, clear plain. In this rather unlikely circumstance, the status quo isn't too bad. There isn't much opportunity for randomness. The character with the higher movement rate should pull ahead of the other; if the predator is faster than the prey, he'll catch him, and if the prey is faster, he'll get away, assuming he can find someplace to hide once he's out of the predator's line of sight. If there's no way to hide, then things are a little more interesting: the character with more staying power (i.e., Constitution) will win out. In other words, in a sprint, the faster character wins; in a marathon, the hardier character wins.
2. The non-trivial chase. Anything more interesting than the trivial chase will involve some sort of obstacle, whether it's difficult terrain, or people crowding a street, or the twisting hallways of a labyrinth. A chase of any decent length will involve a number of different obstacles, possibly calling for multiple checks by the predator and prey. Any number of different skills or abilities might come in handy here.

Okay. So, first things first: character speed shouldn't be the end of the process of determining who wins a chase, but it should be the beginning. Faster characters should have a natural edge over slower ones. Let's turn it from an absolute into a mere advantage:

For every five feet of speed a creature has greater than 30 ft., it receives a +2 bonus to its chase check. For every five feet of speed a creature has less than 30 ft., it receives a -2 penalty to its check.

So, our unencumbered halfling (speed 20 ft.) gets a -4 penalty to his roll, while the human (30 ft.) gets no modifier. The +/-2 per 5 ft. modifier can be tweaked to widen or narrow the chase-speed difference between "fast" and "slow" characters, but let's keep it as it is for now.

Now, what about allowing skills and abilities to help a character in a (non-trivial) chase? I can think of several different skills that could be useful, right off the top of my head, and there are almost certainly some I haven't thought of. So, rather than create a firm list of allowed skills, why not leave it open? As long as the DM doesn't reject it out of implausibility, anything goes. A few examples:

1. Climb. The prey is taking to the rooftops!
2. Jump. The mercenary is in hot pursuit, leaping over crates and wheelbarrows.
3. Balance. The thief scuttles across the rickety old rope bridge with ease, evading capture.
4. Knowledge (local). Here's a more interesting one -- you use your knowledge of the city to cut your prey off at the pass!

Also, ability scores are fair game. A simple Dexterity check will suffice, if no other skill seems more appropriate.

How do we deal with the opposed check itself? What skill opposes what skill?

For the most part, I'd say any skill at all can be used, and the skill used by the prey can be independent of the skill used by the predator, and vice-versa. The assassin can use her Climb skill to take to the rooftops, while the guardsman chasing her uses Knowledge (local) to cut her off. If a character uses a skill that is normally opposed by another specific skill -- say, Hide vs. Spot -- I'm inclined to ignore that normally-opposed skill. If the prey uses Hide to get away, the predator can still use Jump to chase him.

Both parties roll simultaneously. The result leaves much to be decided by the DM (or the players, depending on your preferences); I think that's fine. The rules don't have to tell you if the guardsman catches the assassin before she scales the wall of a building or afterward; the important thing is that he catches her.

An example:

Ander, a human fighter, notices too late a thief stealing a gem from his belt pouch. The thief bolts through the crowded street! Ander vaults onto his light warhorse and takes off after him. Ander uses his horse's speed in this case, which is 40 ft. (modified down from its base of 60 ft. due to medium barding); the thief, a half-elf in leather armor, has a speed of 30 ft. The half-elf uses his knowledge of the city's layout to try to lose Ander, while Ander relies on the speed and power of his horse to catch up to the thief.

Ander rolls 1d20 and adds his speed modifier (+4) and his Ride modifier (in his case, +6), for a total of +10. The thief rolls 1d20 and adds his speed modifier (+0) and his Knowledge (local) modifier (say, +4), for a total of +4. Smart money is on Ander, but it's by no means certain!

The thief rolls a 16+4=20. A tough number to beat, but Ander's roll of 12+10=22 is enough to do it. The thief is nearly run down by Ander's horse a split-second before he can duck into an alley and run to safety; Ander has caught him.

You might also set up a multi-stage chase, in which several opposed checks must be made. I'm not sure how to go about this. Should we mandate that each set of checks be different, with the same skill not being used more than once? How do we (fairly) determine how many stages to use? At any rate, the number of stages must be determined before the chase, so as to avoid the old "roll until you fail" technique used by certain misguided DMs over the years.

Anyway, this is all very stream-of-consciousness, and I'm sure I've missed a few things. Please let me know what you think, and if you have any ideas on how to improve on these chase mechanics, don't hesitate to share them! Thanks. :)

Variant Rule: Automatic Miss Confirmation

It's always bugged me that all characters have at least a 5% chance to miss on any given attack roll. Worse, the more skilled a character is, the more attacks he has per round; the more attacks he has per round, the greater the chance he'll mess at least one of them up.

The DMG proposes a variant rule that gives an attacker a bonus of +10 on a natural 20, and a penalty of -10 on a natural 1. This bugs me, too, having such a steady distribution between 2 and 19, and then blowing up on a 1 or 20.

Actually, I'm cool with the automatic hit on a 20, but wouldn't mind a better solution to the automatic miss problem than the DMG's variant rule. This got me thinking, and I've come up with what may be the foundation for a decent rule. I'm sure it will need some tweaking, but here's my initial thought:

VARIANT: AUTOMATIC MISSES

     When you make an attack roll and get a natural 1, immediately make another attack roll with the same modifier, but also apply a -10 penalty. On this second attack roll, treat a natural 1 as an automatic failure. If the second attack roll is successful, treat your original roll as a 1, and apply any attack roll modifiers. If the second roll is a failure, you have missed your opponent.
     Example: Bob's character, Tordek, has an attack bonus of +16. He tries to attack an orc with an armor class of 13. Bob rolls his d20, and gets a natural 1, so he makes a second roll. On this second roll, Bob gets a 9, which is a success (9+16-10=15 vs. AC 13), so his original attack roll stands as a 1. Applying his modifiers, his final attack roll result is 17 (1+16=17 vs. AC 13), so Tordek hits the orc.
     If Bob had rolled, say, a 6 on his second attack roll, the result would have been a failure (6+16-10=12 vs. AC 13), and Tordek would have missed the orc.

I think the penalty to the second attack roll may need to be tweaked, and, further, may need to be variable rather than fixed. Any thoughts?