Saturday, June 6, 2020

In Praise of the Humble Experience Point

America is having a tough week.  Actually it's been a tough year.  We've got the COVID, the protests against brutality, the escalating police violence against said protesters.  We've got murder hornets.  We're all personally affected by the stuff going on to one degree or another - here on the east coast, I know many people who have lost loved ones to the virus.  I sat down to write something cranky, but figured we've got enough negativity going around.  Let's talk positively about something I do like - the humble experience point.  And heck, maybe there's a way to make friends with the milestone approaches, too.

I had a terrible experience with 4E back when it was fresh and new, and that's what pushed me and my gaming group back to 1st Edition AD&D. We learned what the OSR folks were up to, and embraced the modern analysis of what made those earlier styles of play so much fun.  In fact I'd say the project of my blog has become how best to run 5E in a style that leverages lessons from the heyday of the OSR and early D&D.  Our weekly Tomb of Annihilation game is really close, but I'm not satisfied with the approach I took to managing experience.  That's a story to tell sometime.  So the larger work continues.

Let's step back and distill the essence of this play style I'm praising.  Those early legendary 1970's dungeon masters ran megadungeons, sprawling multi-level complexes.  Game structures were primarily site-based (dungeons or hexcrawls) and featured exploration as the principle motif.  Whatever story is bolted on top the underlying exploration chassis (such as stopping the rampaging giants, discovering the secrets of the evil temple, finding the lich's treasure, pursuing the evil Drow to their underground city) is almost secondary to the exploration.  Players are principally engaged with testing their wits against a hostile dungeon full of challenges, collecting experience points, and increasing their power.  This mode of play maximizes the amount of choice and agency to the players.  The players plan what they want to do each session, including resource planning.  (Ideally the referee collects their ideas at the end of the current session to better prepare for next time).  The game needs to telegraph enough information about the relative risk and reward opportunities so the players can incorporate that into their planning.  This is simple in dungeons, where each new dungeon level down has more dangerous monsters and more wealth.  In the hex crawl, distance from civilization is usually the barometer of danger.

Experience points are a complimentary game mechanic to site-based adventures.  They let the players keep score on how well they're doing in the game.  They're earned, not awarded.  The exponential nature of those early experience charts motivate the players to seek out greater challenges to maintain the same upward momentum.  Note that XP for Gold yields slightly different results than 5E's approach, XP for Fighting.  XP for Gold is an abstraction - all the effort that went into finding treasure - fighting monsters, casting spells, disarming traps, solving puzzles, and so forth, are all assumed to be part of the effort of recovering the treasure.  It's not meant to be realistic, but it is simple, transparent, player facing, easy to track, and non-arbitrary.  XP for monsters defeated isn't horrible, but it does emphasize different behavior.  I've found XP for Gold encourages craftier play, and games better reflect the Sword & Sorcery roots of D&D's earliest literary influences - Conan, Lankhmar, The Dying Earth, those types of tales.

With site-based adventures, the referee can mostly dispense with level-appropriate game balance.  The ref might populate the local area with the goblin mines, the ogre caves, the vampire's tower, and the dragon's lair in the distant mountains.  Or if the principal locale is a sprawling dungeon, you have level 1, level 2, level 3 of the dungeon, and so on.  It's important to telegraph to the players, through rumors, talking to people in the setting, and similar information gathering, which adventure opportunities are going to be more dangerous. The players choose what to go after - it's on them if their first adventure is to go knock on that vampire's tower door.  The creation of these sandbox locales or dungeon levels is really about seeding the setting with experience point opportunities.  It's both art and science calculating how much of a dungeon level or wilderness area you expect a party to encounter before heading on, and populating it with appropriate experience opportunities.

There are challenges with sandbox creation.  First, they can seem like a daunting amount of work - I think of prep time in terms of the sandbox triangle (you can have a lot of detail, but it takes a lot of work; or you can build out a bigger area with not a lot of detail for the same time investment).  Older editions put a lot of emphasis on random tables, both wilderness encounters and dungeon wandering monsters, to create a sense of a living world and give the referee some help creating content on the fly.  Finally, while there will be story reasons for various lairs and dungeons in the setting, and "plot hooks" that may motivate the players to go explore them, sandbox games are less about scripting an intricate story-line in advance, instead turning the keys over to the players and seeing what emerges from their activities.

I love site-based settings and exploration based play, and I strive to turn as much of the decision making and planning over to the players.  Incidentally, these are still the most popular adventure styles with the 5E crowd, too; go to any ranking list on the official 5E adventures and products like Curse of Strahd, Tomb of Annihilation, or Lost Mine of Phandelver, are consistently top of the lists, along with Tales from the Yawning Portal or Ghosts of Saltmarsh.  Hint:  they all feature exploration-based dungeons and open world sandboxes.  I've been wondering how it would look to shift from having the players bean-count their experience points to using something even more abstract like milestones.  Below are a few recent attempts.

Dungeon of the Mad Mage
When I was running Dungeon of the Mad Mage, the gigantic 23 level megadungeon for Waterdeep, I dispensed with experience points.  Mad Mage's levels are calibrated to where a 4 person party needs to literally clear (as in fight, kill, or drive off) every single monster on a given level in order to collect enough experience to level up.  It's tiresome to even think about, and way too much of a slog to be enjoyable.  Nope.  Instead I made the discovery of each new dungeon level into a milestone - the idea being the effort to explore a sprawling dungeon level, overcome traps, challenges, monsters, and so forth, represented achievements worthy of advancement (either a full or half level gained).  Mad Mage's staircases are geographically remote on each dungeon level, requiring a party to negotiate large swaths of the dungeon level before descending.  Normally I'd consider that a poorly designed map, but in this case those remote stairs became a feature, supporting exploration-based milestones.  That campaign went on the shelf due to COVID, so we only got through the first few levels, but it was going exceedingly well.  The players focused a lot more on scouting, negotiating with monsters, and using wits and guile to find those staircases in lieu of slaughtering every last monster.  In this case, milestones worked well as stand-ins for experience points - they were player-facing, transparent, allowed the players to keep score, and influenced player planning.

Mad Mage is not popular with the vocal part of the 5E crowd.  People look at 23 dungeon levels without an overarching scripted story, and they don't know what to do with it.  I'll add a discussion of Undermountain to my backlog on posts I'll get to at some point - what we did to make it more engaging.  Its not hard, but the dungeon master does have to do some work to overlay interesting story goals onto the megadungeon.

Dragon of Icespire Peak
Icespire Peak is the starter adventure in the second 5E boxed set (the Essentials Kit).  On it's face, it describes a sandbox type area of the Sword Coast, with 12-14 adventure sites.  There's a loose story in the sandbox - a white dragon has moved into the nearby area, and this has created some ripple effects that have put the sandbox in motion.  The dragon has displaced a mountain orc tribe, and the mountain orcs have descended into the valley, attacking places or displacing other monsters that are now encroaching on the villagers.  It uses a quasi-milestone approach... gain a level for each starter lair completed, then gain a level when completing two mid-tier sites, and so on.  I'm running a new Tuesday night game with some of the adventurer's league guys via Zoom, bi-weekly, so we're getting some experience with this one.  The adventures are presented as quests from the town master; the next 1-2 quests become available as the players finish the prior ones and level up.  The fetch-quest approach isn't awful; the players have been able to collect a couple of quest ideas at a time from within town, map them out, and plan efficient ways to go tackle exploring 1-2 locales on an excursion out into the wilds.  It's still enabling player-facing planning and decision making.

The village and Town Master is lackluster, and I'm finding it's critical there are interesting and engaging NPC's so the players learn more about the sandbox region.  There are cool places on the map to explore, not tied to any quests, and the players need to hear about them from NPCs.  As written, the quests and locales don't telegraph to the players the level of danger at each site.  That's an element the referee needs to work into the player-facing aspect of Icespire Peak.

Neither of these approaches to merging milestones and site-based adventures left me completely satisfied.  I suppose the Mad Mage approach was closest.  Listing out the attributes I like about experience points - they're simple, transparent, player-facing, easy to track, and objective (ie, non-arbitrary) - the Mad Mage approach comes nearest to meeting the requirements.  Unfortunately it puts a heavy constraint on how you construct your dungeon maps, and doesn't translate equally well to lairs or the hexcrawl space.  Might just be easier to maintain using experience points, as they apply equally well in most situations.  Would love to hear if any readers have successfully ported milestones into their exploration-based dungeons.


  1. It sounds as if I'd like to play with you! Site-based settings with an emphasis on exploration and player choice of direction--sign me up! It's the kind of game I try to run, too.

    Yet I do without experience points in the games I run. I appreciate the goal of having a seemingly non-arbitrary and trackable metric for advancement, but I think that, while points *are* trackable, they are ultimately arbitrary. The dungeon (or scenario) design has already been programmed or seeded with potential experience points to begin with, so those humble and pleasant experience points are basically power-up points waiting to be discovered. The arbitrariness of them is less visible behind a layer of pre-programming. It's prestidigitation: XP *seem* earned but they were actually placed there beforehand. Their placement makes up the milestones. Milestones without XP are just as pre-programmed, in my view, so I just say, what the heck, level-ups come at the intervals that I decide. Players seem to know when they are due, too.

    In the end, whatever is fun is best, right?

    1. I like the idea behind milestones - simplifying the accounting side of advancement. But the primary value of XP is their transparency and the drama it creates for the players. I'd need to find a good way to duplicate that quality with milestones in a sandbox game.

      Here's an example from a recent Tomb of Annihilation game with my home crew. They were defeated after a grueling duel with a beholder, retreating after losing several party members. The beholder was guarding a significant horde of treasure and magic. But defeating the Beholder was not germane to their current goals, it's a side trek. As a team, they had a vigorous debate whether it was better to revisit the Beholder, to avenge their comrades and reap the experience benefits, or let it go and carry on with their actual goals. They passed on fighting the beholder a second time and prioritized the story goals. The debate wouldn't have had nearly as much drama if experience points weren't sitting on one side of the equation. Someone could certainly argue for the other point of view! Ie, "We shouldn't lose out because story is more important to us than personal advancement".

    2. XP for GP is a lot less arbitrary than milestones, Lich. As a DM, yes, I decide where the monsters and treasures go. But the players get to decide where to go, which monsters to seek out, and which treasures they want to try and gather. So it's more of a long as I as DM give the players some ideas of the risk/rewards, and the freedom to choose where they want to go.

      If you've just got the players on a railroad narrative (not accusing you personally of this, using the general "you" here) then yes, treasure rewards and milestone rewards are equally arbitrary.

  2. An excellent summary and analysis of the nature of experience points and how classic experience points differ from modern experience points. A great read, too!

  3. I'm about to start running Curse of Strahd. The book encourages milestone leveling, where the DM just announces that everyone goes up a level when he or she feels it appropriate (after significant story elements are accomplished). My problem with this is that it gives the players no way to track their progress. Even though XP have always been in some sense arbitrary, as a player I like being able to see how close I am to the next level, and it's even better if I know what I can do to move that forward faster (that's why XP for GP is so effective). As a player I don't really like "level up when the DM says so", because there's no sense of any accomplishment on my part.

    So for CoS, I've decided to go with traditional XP for defeating monsters (not necessarily killing, but just resolving the threat), augmented with XP for major and minor milestones like the DMG suggests. So significant accomplishments will be worth a hard encounter's worth of XP, and minor goals (perhaps individual PC goals determined by the players) will be worth an easy encounter. I think if I can get these goals clearly articulated as a sort of rolling list, all the better, because the players will be able to see very plainly what they can do to earn XP.

    I'd be interesting in hearing how your Curse of Strahd sessions went and what you did for XP.

    1. I think the challenge for something like Curse of Strahd will be tallying up the likely encounters and then verifying if the players can actually gain close to enough experience to match the milestone pace. Whatever the gap is, that's what you'd need to supplement with quest goals. I like the approach of making the quest goals public (as they become revealed) so they fill a similar role to XP in motivating the players. It's a subtle shift, but completing a quest feels like the players earned it, as opposed to the players doing something, and then after the fact the referee mentions "here's some XP for that thing you did". Mechanically they're the same, the impact on play is quite different.

      Going off memory on CoS, the various items in the Tarot deck (um, Tarroka deck) could all have major XP awards tied to their discovery, and then fill in the gaps as smaller quests come up, like getting Ireena someplace safe out of Barovia, which becomes a potential quest very early. CoS may be on our horizon, as we wrap Tomb of Annihilation I know at least 1 of my players is a relentless agitator for Barovia. "I have a burning fever, and the only cure is more Strahd."

  4. Everytime I see that image at the top of your post, I think of Jeff Rient's blog. He had that as his banner for the longest time.