Evasion, Flight, and Pursuit – Chases in TTRPGs

I’ve been in a conversation about the encounter evasion mechanics in OD&D (Vol. 3, pp. 19-20) on the Wandering DMs Discord server, and I thought it’d be worthwhile to go through those rules here, along with further musings on the topic.

To start out, the book presents a roll to determine if castle occupants that emerge to confront a party will pursue them. It’s a simple and straightforward roll, with the odds depending on whether the occupants’ reaction was hostile or merely neutral towards the party. I wish there was some further consideration about the size of the castle’s territory of control, but that’s easy enough to make up if I need something. So far, so good.

The first real mess comes up immediately after that. There’s a table for the chances of evading the encounter based on the size of the party, the percentage of the maximum possible number of opponents encountered, and footnotes providing modifiers for surprise, terrain type, and whether one side has higher travel speed than the other.

Party size is reasonable variable; it might be too fiddly for many younger groups, but since this book was written for 4-50 players who were expected to employ hirelings in addition to their PCs, I don’t mind treating a party of five differently from a party of thirty.

Using the proportion of opponents encountered compared to their maximum instead of absolute numbers, on the other hand, just seems bizarre. For one thing, the text isn’t entirely clear on what “possible #” refers to; the maximum of the “Number Appearing” column in the Vol. 2 monster table seems like a reasonable baseline, but then do you also include the extra leadership/elite positions that come into play in their descriptions? In the case of castle occupants (or a mixed wilderness encounter, though that seemingly wasn’t a consideration at the time), do you track each group separately or consider them all collectively? What kind of sense does it make for two dragons to be 40-400% easier to escape from than one dragon? I can only guess that using a proportion instead of an absolute number was intended to normalize for the wide span of encounter sizes (from “1” on the low end to “40-400” on the high), but if the party size counts can treat 10-24 as a single range, it seems like it would’ve been reasonable to have rough counts for the opponents, too (maybe 9 or less/10-24/25-99/100+, for instance, to use similar breakpoints).

There are some oddities in the footnotes, too. Opponents having surprise removes any chance of evasion whereas the party having surprise merely doubles the chance to evade (hardly the only adversarial double-standard in OD&D, admittedly, but I find asymmetric mechanics distasteful outside of storytelling games). Woods are the only terrain with direct impact on the chance to evade (sort of), even managing to give some small chance to evade if surprised. The speed comparison is a single modifier based on which side is faster, no matter if that’s by 5% or 500%. Also, it’s unclear if the percentages given for terrain and speed are additive (as percentage points) or multiplicative (as an actual percent); I’d guess the former if only because none of the base evasion chances is evenly divisible by four and most aren’t evenly divisible by ten, but given the proportion calculation for the number of opponents, either interpretation is reasonable.

In any case, once the reader sorts all that out, it’s on to the actual pursuit procedure (note that there’s quite some ambiguity in how the actual text can be parsed, so the steps below are just my best attempt to make sense of it):

1. Roll a die to determine a random direction to move.

2. Check if pursuers with a speed advantage catch the party. If successful, transition to combat.

3. Move to the next hex (based on the direction rolled in step 1).

4. Check the chance of evading. If successful, the party escapes.

5. Repeat from step 1 “…until pursuit is ended or melee occurs.”

6. Oh, and add a bonus to the chance to evade in swamp terrain because editing was not a strong point in these books.

There are a few things to unpack here.

First, castle occupants apparently give their full commitment when they decide to chase someone, because the text is explicit about separating their roll to pursue from the rest of the evasion/pursuit rules. I’d personally keep it in there (if I wasn’t going to just handle it by fiat), if only because the alternative seems rather dumb. Also, it’s possible to read the text to say that castle occupants only roll to pursue while other opponents only trigger evasion checks, but this post will bloat even more if I pick at all the semantic ambiguities.

Second, faster pursuers are the only ones who get an explicit roll to catch the party. Same speed or slower? It seems like it’s left entirely up to the party to decide if they continue to flee or engage in melee. Whether slower pursuers matter is debatable (I think they should have chances to catch up if they could reasonably close the separation over long distances, e.g. zombies or robots that don’t need any rest or sleep, or if they could herd the party into terrain where they become faster), but having no mechanism for pursuers of equal speed is odd.

Third, there might actually be such a mechanism, though. This is more likely than not excluded by a strict reading of the rules as written, but if you treat the pursuit as both sides entering the same hex, it could be argued that there should actually be another surprise check in between steps 3 and 4. I’d reduce the chance for surprise, unless the party was explicit about not being wary of continued pursuit, but I think keeping that chance in there would be a nice mechanism to add uncertainty.

Fourth, if the previous idea about rerolling surprise is ignored, there’s no clear way to end a pursuit by opponents who aren’t faster except when the party chooses so. The “…until pursuit is ended or melee occurs” condition just kind of dangles without resolution. While there are a few indirect limits on how long the party can choose to flee (food, water, chance for wandering monsters, fear of getting lost, potentially the map limits if you’re using the actual Outdoor Survival board), there’s no direct check for if the opponents catch the party unless step 2 is in play. I think there ought to be some chance, since not having any gives the party too much control over the situation.

Finally, only melee matters for ending a pursuit in the rules as written. There’s no suggestion for how to handle bows or spells, which seems like a questionable oversight because there’s no reason why pursuers should have to follow slasher villain logic. It wouldn’t be difficult to let the pursuing side choose if they want to engage from range (which would risk giving the party a chance to flee successfully) or close to melee (with the obvious risks that could bring), but then that would beg the question of where it would fit in. If it happens when the pursuers catch the party, there’s arguably no real choice being made. If it happens either immediately before or after moving the party to a random hex, how much impact should it have on the chance to evade in step 4?

For how little space the book spends on this final paragraph, it’s impressive that it introduces so many questions.

After that procedure, there’s a tidbit that each hex moved during a pursuit requires half a day of rest and that wandering monster checks during the resulting full days of rest have a base 2-in-6 chance instead of the typical 1-in-6. That seems straightforward, until it comes to the matter of when those half-days happen. If it’s during the pursuit (which is what feels intuitive to me from a simulationist standpoint), then it’s hard to say what effect they actually have since both sides would suffer that penalty (zombies/robots/etc. notwithstanding) and the doubled chance for wandering monsters would seemingly never come into play. Also, would this mean the pursuit moves only three hexes per two days (base foot speed of 3 hexes per day, less 1.5 hexes for the half-day of rest), or are we to assume some balance of faster movement when they’re moving with longer breaks to rest from moving results in the same average movement speed as just traveling normally (which seems silly)? If it’s after the pursuit (and potential ensuing combat) ends, that would resolve bringing the wandering monster checks into play, but it results in a logic gap where the party can run forever but then needs multiple days to recover as soon as the chase ends, no matter if that takes four hours or four weeks.

With all of that having been said, I can get to my thoughts on the matter.

First, let me give Arneson and Gygax credit for including an attempt at pursuit rules. Despite the rough spots and incompleteness, just having the idea in the rules is valuable for getting players (on both sides of the proverbial screen) to think about flight as a valid alternative to combat, which leads to interesting choices (similar to the rules for subduing dragons introducing the seeds for considering surrender or taking captives as valid end states for combat). Regardless of the actual mechanics, the game experience is richer for the idea.

Now, what are chases trying to accomplish? Evasion/flight signals that one side of a combat feels the potential rewards are not worth the risks, and pursuit signals that the other side feels the potential rewards are worth the extra effort to prevent their foes’ escape. On a simple mechanics level, there needs to be a way to resolve one side or the other getting what they want. Whether to handle it as a single-step zero-sum check or an extended series of checks is a matter of taste. I like drawing out tension in chases and giving both sides opportunities to influence the outcome since I view chases as filling a similar role to combat (in terms of how they each affect the player’s experience), but that’s not for everyone.

Since this is my blog, though, I’ll talk about what I want from a good chase mechanic.

On an intuitive level, chases should feel frantic and intense, regardless of whether they’re happening in a matter of seconds (in a dungeon) or days (in wilderness). That means they should have very quick resolution for each individual roll (minimum necessary modifiers, no table look-ups) and multiple rolls between any references to maps.

Next, everyone involved in the chase should have opportunities to influence the outcome. Chases where people try to do stuff are simply going to be more interesting than chases where everyone just runs. In keeping with the previous point, I’d want to alternate who has a chance to make an impact between each side, since that back-and-forth feel should maintain tension better than having everyone on one side act before the other (particularly with the condition in the next paragraph).

Third, the chase mechanic needs the flexibility to work with both small and large numbers on either side of the chase. Even if there are asymmetric mechanics between PCs and NPCs (which I’d dislike, as I’ve said, but I know plenty of people want PCs to be special on a mechanics level in TTRPGs), the mechanics should be able to handle fleeing from or pursuing a powerful single entity, a rough match to the PCs in numbers/power, or a horde of weak opponents reliant on strength of numbers.

Fourth, as with any mechanics, it should be something that gets satisfying and sensible results without sacrificing or interrupting the feel of the game. Since TTRPGs tend to have tons of rules, both codified and made up on-the-fly in play, each individual rule/subsystem should be as easy to get into and out of as possible. Fortunately, the needs here overlap with much of what helps to keep a frantic and intense feeling, with the extra caveat that it should be easy for a player who doesn’t understand the full details of the chase mechanic to intuit what they can do to influence the outcome (per the second requirement) with minimal explanation.

Lastly, chases should be able to suit two different sets of conditions: chases that can continue indefinitely until one side wins and chases that are more like races in that one side is trying to prevent the other from reaching a goal (e.g. the PCs are chasing the cultists to stop them from reaching the ritual site in time to summon the Dread Lord Vomitus). If the chase mechanics allow for easy transition from one to the other (e.g. having failed summon the Dread Lord Vomitus, the cultists turn to just trying to escape), that’s even better, but I see that as more of a stretch goal.

Having laid out what I want, that begs the question of whether I’ve found it. I have, to a degree. In games that lack these mechanics (or in which they’re implemented poorly/incompletely, like OD&D), I take care of chases with my general minigame for things that don’t have formal subsystems but would benefit from extended resolution. I’d like something more specialized that is built to really capture the feel of a chase, but while I’ve heard or read about some interesting ideas, I haven’t yet played with anything that met my desires better than my general minigame.

This has gone on longer than I’d intended to, so I’ll save actually talking about my general minigame and how I use it do run chases for a follow-up post. For now, though, if any readers have thoughts or suggestions on chases in TTRPGs, please share them.