2026-10-01

Exploration XP Complexity

As any economist will tell you, people respond to rewards.  So reward the behavior you want.  As a game designer or GM, XP are one of the primary rewards, so you should carefully choose what to award them for.  Don't want people murdering baby kobolds?  Then don't give XP for murdering baby kobolds.  Don't want murderhobo behavior in general?  Don't give XP for killing anything.  Don't want loot goblins?  Don't give XP for treasure.  Want PCs to explore more?  Give XP for exploration.  Want them to solve your precious little story quests rather than thinking up goals for themselves?  Only give XP for solving your quests, nothing else.  (Train whistle sound effect.)

(There is, as always, a tradeoff between not railroading and not letting the campaign devolve into "let's do nothing and get free XP."  Ask any manager who let their employees set their own goals, there's always that one guy with the goal of doing no work and getting promoted for it.  So you have to have some standards for what's XP-worthy, to keep the game interesting.  In DF Whiterock I got kinda lazy and just handed out fixed XP for showing up, and the players responded by exploring as slowly as possible and never taking a risk and then getting overpowered for the level they were on to the point where the dungeon was no longer a challenge and then whining when I reduced XP awards.  So, again, award XP to encourage the behavior you want, but in a way that hopefully intersects with behavior the players are also okay with.)

Exploration XP is pretty popular among both players and GMs.  Players like to explore (though perhaps maybe not as boldly as the GM would like), and GMs want players to actually engage with more of the adventure, so it's a win for everyone.  In DFRPG Arden Vul it's one of the two main sources of XP, along with loot.  (Yes, my players are loot goblins.  You should see their spreadsheet.)  Exploration XP is easier to calculate than loot XP though.  You just decide on what counts as a new location, what counts as exploring, and then count how many new locations they explored each session, and then give the appropriate amount of XP for that number of locations.

GURPS/DFRPG has "big" XP, where 1 XP is enough to buy one level in a new skill.  Improving a skill may take up to 4 XP, and buying a new advantage might be 5 or 10 or 15 or even more.  But contrast with games with "small" XP where you need thousands of XP to "level up" and get a prefab pile of bonuses.  So, basically, I'm awarding 1 or 2 XP at a time, not hundreds or thousands.  Human-scale, not pinball-score-scale.

Here's the chart I've been using for DFRPG Arden Vul:

New locations XP
0-4 0
5-19 1
20+ 2


Pretty simple.  The first question is what counts as a location.  I've mostly just deferred to the Arden Vul adventure.  The adventure has keyed room locations, so I use those.  Sometimes there are edge cases, like there might be a 1-1A and a 1-1B right on top of each other and I might decide that together they only count as one location.  But for the most part, the locations on the map are the locations I track.  And I use the actual room numbers from the adventure rather than adding my own set of labels.

The second question is what counts as exploration.  This is trickier.  If the PCs merely glimpse the location from far away (like with scrying), that's not good enough.  I've seen Los Angeles on TV many times, but I've never actually been to LA.  No XP for me.  So the first threshold is that they need to actually personally enter the location.  No crystal balls, no Wizard Eyes, no possessed scout bats.  (Those are all useful for gathering information, but they don't count for XP.)

Now, does just running through count?  I've changed planes in Detroit Airport, but does that mean I've actually been to Detroit?  I think not.  And I'm not talking about technicalities of whether the airport is actually within city limits.  I mean, if you've only seen the inside of the airport, you haven't actually seen enough of the city to tell it apart from the hundreds of other cities with large airports.  Exploration means actually exploring an area, not just sticking a toe into it.  You should have heard the full room description and probably interacted with most of the interesting things in the room.

Now, some rooms are just barren, so no matter how much you interact with them you don't get much.  There aren't a lot of those in Arden Vul; modern dungeon design (even modern dungeon design for ancient games like AD&D 1E) discourages too many empty rooms.  If you've found whatever there is to find, sure, you've explored enough.

On the other hand, some rooms are complicated.  They contain monsters and treasure and secrets and features to interact with.  Do you have to completely experience everything in the room to have explored it?  I don't think so.  But you need to have at least interacted with the room a reasonable amount.  Maybe you haven't fully solved the 13 clocks puzzle, but you need to have at least noticed that there were 13 clocks.  If you were too busy fleeing from rabid weasels to see the clocks, you haven't actually explored that room yet.  Come back later after you've had your shots.

Finally, I'll mention long-term tracking of explored rooms over the course of a long campaign.  A feature of a megadungeon is that PCs tend to revisit the same areas over and over.  But you don't want to credit them for exploring the same room more than once, or they'll notice and stop exploring and start hiking back and forth over safe territory for XP instead.  As the PCs explore, I note each room they explore in a text file, along with other notes like who they meet or what they take.  I have a simple Python script containing a dictionary of session number to set of room numbers explored, and after each session I update the dictionary with the explored locations for that session.  Something like:

"s1": {
        "EX-2",
        "EX-8",
        "EX-9",
        "EX-10",
        "AV-1",
        "AV-5",
        "AV-6",
        "AV-7",
        "AV-8",
        "AV-14",
        "AV-15",
        "TS-12",
        "TS-13",
    },

So in session 1 ("s1"), the PCs explored 13 locations, good for 1 exploration XP.

The value of using code for this rather than just tracking on a piece of paper is that the script puts all the previously explored rooms into a set, to avoid double counting.  So if they visit TS-12 again in session 45 and I've forgotten they've already explored it, the script should avoid giving double credit for the location.


No comments:

Post a Comment

Exploration XP Complexity

As any economist will tell you, people respond to rewards.  So reward the behavior you want.  As a game designer or GM, XP are one of the pr...