The Alexandrian

Revisiting Encounter Design

August 30th, 2008

One of the first reviews I ever received for a book I had written was for the mini-adventure The Dragon’s Wish, which was published by Fantasy Flight Games during the early D20 boom. The reviewer hated it. He had several reasons for doing so, but his biggest problem was that he felt that the encounters weren’t balanced: The adventure was designed for 9th level characters, but I had them encountering, among other things, a primitive tribe of kobolds (low CR) and a pair of extremely powerful stone golems (high CR).

Now, The Dragon’s Wish was one of my first published works and it was hardly perfect by any stretch of the imagination. But there were two main reactions that I had to the review.

First, I was frustrated because the reviewer had clearly failed to understand what the adventure was about. He had approached it as some sort of hack ‘n slash affair, but the module wasn’t designed with combat in mind. At the beginning of The Dragon’s Wish, the PCs are asked by a dying dragon to take his heart to the ancient draconic burial grounds in the Valley of the Dragons. The rest of the adventure is a travelogue allowing the PCs to see various facets of draconic mythology. The stone golems aren’t meant to be fought: They were powerful gatekeepers who allow the PCs to enter the valley when their task is made known. The kobolds are a primitve tribe who venerate the dragons without truly understanding them. And so forth.

Second, I realized that something fundamental had shifted in the common perception of what constituted proper encounter design in D&D.

Back in the halcyon, nostalgia-tinged days of 1st Edition, nobody would have blinked twice at the idea of including low-level encounters in high-level adventures. For example, in the Bloodstone modules (the original H-series designed for levels 15 thru 100), the designers had no problem including combat encounters with common orcs.

In fact, this was an attitude that persisted more or less all the way through the latter days of 2nd Edition. The Apocalypse Stone was a high-level adventure published to provide a campaign-ending scenario so that groups could reboot fresh with 3rd Edition. But if you flip through it, you discover quite a few encounters that are virtually identical to the types of encounters found in low- or mid-level modules. (There’s harder stuff too, of course.)


So what happened in 3rd Edition?

As far as I can tell, everybody misread the rulebook. Here’s what the 3rd Edition Dungeon Master’s Guide had to say about “Encounters and Challenge Ratings” (pg. 100):

A monster’s Challenge Rating (CR) tells you the level of the party for which the monster is a good challenge. A monster of CR 5 is an appropriate challenge for four 5th-level characters. If the characters are higher level than the monster, they get fewer XP because the monster should be easier to defeat. Likewise, if the party level [….] is lower than the monster’s Challenge Rating, the PCs get a greater reward.

And a little later it answered the question “What’s Challenging?” (pg. 101):

Since every game session probably includes many encounters, you don’t want to make every encounter one that taxes the PCs to their limits. They would have to stop the adventure and rest for an extensive period after every fight, and that slows down the game. An encounter with an Encounter Level (EL) equal to the PCs’ level is one that should expend about 20% of their resources — hit points, spells, magic item uses, etc. This means, on average, that after about four encounters of the party’s level the PCs need to rest, heal, and regain their spells. A fifth encounter would probably wipe them out.

And, at that point, everybody apparently stopped reading. Because this was what seeped into the collective wisdom of the gaming community: Every encounter should have an EL equal to the party’s level and the party should have four encounters per day.

I literally can’t understand how this happened, because the very next paragraph read:

The PCs should be able to take on many more encounters lower than their level but fewer encounters with Encounter Levels higher than their party level. As a general rule, if the EL is two lower than the party’s level, the PCs should be able to take on twice as many encounters before having to stop and rest. Two levels below that, and the number of encounters they can cope with doubles again, and so on.

And if that wasn’t clear enough in saying that the PCs should be facing a wide variety of ELs, the very next page had a chart on it that said 30% of the encounters in an adventure should have an EL lower than the PCs’ level; 50% should have an EL equal to the PCs’ level; 15% should have an EL 1 to 4 higher than the PCs’ level; and 5% should have an EL 5+ higher than the PCs’ level.

But all of that was ignored and the completely erroneous “common wisdom” of “four encounters per day with an EL equal to the party’s level” became the meme of the land.

By the time The Forge of Fury was released as part of the original Adventure Path in late 2000, the meme had already taken hold. The Forge of Fury — an adventure for 3rd to 5th level characters — included, as one of its encounters, a CR 10 roper. You’ll note that this encounter follows the guidelines printed in the DMG precisely. It didn’t matter. The fanboys howled from one side of the Internet to the other about this horrible and unbalanced encounter. And why were they howling? Because encounters should always have an EL equal to the average level of the PCs.

WotC never made that “mistake” again.


The most virulent form of the meme was rarely followed in its strictest form. But the general meme of “an encounter should almost always have an EL equal to the party’s level” sunk pretty deeply into the collective consciousness.

But there are consequences for designing encounters like that:

(1) The average resolution time for any combat encounter increases (because a more challenging opponent takes longer to overcome).

(2) The PCs are more likely to suffer grievous injury during any one encounter, which means they’re more likely to adopt cautious styles of gameplay. This leads to the 15-minute adventuring day becoming more common, along with all the problems that creates.

(3) These factors result in fewer encounters during each game session, which means that it becomes much more difficult and/or tedious to run the classic mega-dungeons and other combat-oriented styles of play.

(4) The utility of any given monster is significantly reduced because the range of levels in which you can build “appropriate” encounters using the creature is narrowed.

I used to play D&D with my friends during lunch hour, and in these short sessions we would still routinely get through 3 or 4 combat encounters. But with 3rd Edition people were routinely reporting relatively simple encounters taking hours to resolve.

A lot of people blame the system for that. But, in my experience, it’s all about the encounter design.


When I looked at the design of classic modules from the ’70s and ’80s, I discovered that most of the encounters in those modules would actually equate to an EL at least 2-5 levels lower than the party. And when I duplicated that encounter design in 3rd Edition, combat predictably speeded up.

With that in mind, here are my tips for designing encounters:

(1) Design most encounters around an EL 2 to 4 lower than the party’s level.

(2) Don’t be afraid of large mobs (10+ creatures) with a total EL equal to the PCs’ level. The common design wisdom is that these creatures are “too easy” for the PCs. This is true if you’re thinking in terms of the “common wisdom” that sprang up around misreading the DMG, but in practice these types of encounters work just fine if you’re looking for fast encounters and lots of them.

(3) Encounters with an EL equal to the PCs’ level should be used sparingly. They should be thought of as “major encounters” — the memorable set pieces of the adventure. It actually won’t take very long before the expectations of your players’ have been re-aligned and these encounters leave them thinking, “Wow! That was a tough encounter!”

(4) And that means you get even more bang for your buck when you roll out the very rare EL+2 or EL+4 encounter.

Basically what you’re doing is creating a wider dynamic range for your encounter design.


FLEXIBLE DESIGN: I like to design large complexes of opponents who will interact with each other and react, as a group, to the presence of the PCs. And this works a lot better if I can take two encounters and add them together without ending up with something that will completely devastate the party. If the PCs are level 5 and the goblin warband is only EL 3, then it become much easier to have the goblins call on a second warband to reinforce them: If the PCs prevent the reinforcements from showing up, they have two standard encounters. If they don’t, then they have one harder encounter.

EXPERIENCE POINTS: The designers of 3rd Edition increased the pace at which XP was accumulated and levels were gained. I understand and even support the reasons behind this change, but I personally found the result to be simply too fast for my taste. For example, I tend to run long 8-12 hour sessions, and the pace of 3rd Edition experience usually meant that the PCs were leveling up once per session. This meant that the power level of the campaign shifted very rapidly (making it difficult to tell coherent stories). It also meant that the players never really had a chance to get comfortable with their characters (they had barely learned one set of abilities before being given new ones).

I now play with halved XP rewards and have had good results with that. But, really, that’s just a matter of personal taste.

However, with that being said, using the encounter design recommended here, you’ll find that your players will be overcoming many more combat encounters in the course of an average session. And even though the EL of each encounter will be lower, this will still generally result in accelerating the already accelerated pace of XP accrual. Whether you’ll need to adjust the XP award accordingly will depend on your personal tastes.

A PERIOD OF ADJUSTMENT: If your group has already grown accustomed to the “typical” design of 3rd Edition encounters, it may take some time before your expectations have adjusted to the new system. The typical encounter will feel easier to you… and that’s okay. It is easier.

But you should also be aware that some of the secondary effects will also take awhile to sink in for your players. If you’ve been playing with “typical” 3rd Edition encounters, then your players have probably learned to take a very cautious approach — every encounter has been potentially deadly and, therefore, every encounter has been carefully analyzed and handled.

So for the first couple of sessions, for example, you may only see a slight increase in the pace of gameplay. But once your players internalize the change and loosen up, you’ll see that pace increase again.

Pay attention to your own expectations, too: You might find yourself getting a little frustrated with the fact that your villains are missing the PCs more than they’re hitting them. There’s a sense that a lot of us develop that says “if hit points aren’t being lost, then nothing happened”. This isn’t actually true. And, in fact, if the PCs aren’t losing hit points the more stuff will happen.

DIFFERENT TOOLS FOR DIFFERENT JOBS: The exact balance of combat encounters you choose will depend largely on the type of adventure you’re designing. For example, if you’re designing an intrigue-laced adventure in which the only combat encounter is likely to be the big show-down at the end… well, that single encounter should probably be a doozy. If you want to encourage a loose, rapid-fire style of play with the players feeling like major heroes… well, crank up the number of low-EL encounters.

If there’s one message to take away from this essay it’s that variety is the spice of encounter design. By extending the dynamic range of encounters, you’re expanding the variety of the encounters you can (and should) design.

Share on TumblrTweet about this on TwitterShare on StumbleUponShare on FacebookShare on RedditShare on Google+Digg this

16 Responses to “Revisiting Encounter Design”

  1. Justin Alexander says:


    This type of setup is perfectly suited to the Clone Wars era campaign my group is currently running. B-1 droids are hardly a threat, it’s the B2 Supers you have to really watch for. It also allows me to do things that were seen in the movies, such as B2’s firing from the cover of a phalanx of their smaller brethren (Often resulting in casualties to aforementioned Phalanx) The increased number of enemies isn’t a problem either as I treat the B1’s as being networked into a cell of 3 or 4 droids, moving in a somewhat coherent formation and firing at the same target, or if I’m in a hurry having them aid-another each other to represent the fact they’ll hit something through massed fire.

    The players hate running into little squads of B1 droids, as even though they go down fairly quickly, they’ve almost certainly raised the alert. The guys who’s troopers are carrying heavy weapons get a kick out of this too as against a handful of strong enemies ,the AOE of their weapons would usually be wasted. However a cell of 4 Battledroids being vaporised is always pleasing to them. Lets them feel like their getting the most of the weaponry they chose.
    Sunday, February 28, 2010, 8:16:01 AM

    I believe what Roy was referrring to when he said ‘routine encounters’ was CR=CL encounters.
    Wednesday, December 30, 2009, 11:22:18 AM

    Justin Alexander
    Uh… I actually said that routine encounters should be exactly the opposite of that.
    Friday, June 19, 2009, 8:27:51 AM

    Wow. This is a Fail article. Mostly because you think routine encounters are slow and deadly.
    Wednesday, June 17, 2009, 1:18:17 PM

    Good read. Linked here from GitP. One of your better essays in my opinion. Very insightful.
    Saturday, September 06, 2008, 3:48:48 AM

    I never really was too concerned with encounter balance. I knew that as GM I could “fix” anything that went horribly “wrong”. Bowing too deeply at the altar of game balance is never a good thing.
    Tuesday, September 02, 2008, 9:50:28 AM

    Jonathan Drain’s D&D Blog
    +1 Insightful.

    One of my biggest complaints with third edition is that it encourages sleeping after every combat to regain daily abilities. I ranted about this last year:
    Tuesday, September 02, 2008, 3:55:52 AM

    Justin Alexander
    @Andy: That is a very good point. And you’re right, it’s one I should have raised, because it’s something I really love to do when I’m DMing.

    On a small scale, we actually had something like this happen in our last session. The campaign is centered around Ptolus. When the PCs first got to town, they were getting involved with some street punks and the like who were causing problems. For the past several weeks, however, they’ve been struggling against demons and goblins and the like… and they’ve been leveling up. So when they ended up struggling with street punks again, they were first taken aback and then gleeful when they realized that they were push-overs now.

    That sense of transition from common people to heroes to movers-and-shakers is something I really love about D&D.
    Monday, September 01, 2008, 12:36:26 AM

    Interesting read Justin.

    I had a mini version of your “more lower CR fights = faster flowing play” epiphany on reading the “Dungeonomicon” homebrew. The authors of that work openly state that there should be swarms of lower-than-PC-CR creatures cluttering up fights (climactic or otherwise).

    Low CR mooks and followers are there to make the PCs and the Big Evils look awesome. They’re the schmucks that stuff like fireball, flurry of blows, great cleave and so forth are intended to take out, and take out by the handful.

    This ‘king of the hill’ thing can go on right up until the Big Bad (Mr CR+4) storms down the steps of the skull throne roaring “Fools! Weaklings! I’ll dispose of them myself!”, at which point ‘heroic badassness’ takes a back seat and ‘heroism in adversity’ kicks in.
    Sunday, August 31, 2008, 11:45:55 AM

    Andy Patrick
    There’s one more reason for a range of encounter difficulties that you overlooked until one sentence very near the end but that I think is critical.

    If CR = player level throughout, there is never any variation in challenge. So the characters never feel like they are progressing. I call this “Oblivion-itis”.

    Suppose a 1st level party faces off against 5 Kobolds and wins only narrowly. At a higher level, if all they fight is 5 Ogres, the experience will be identical. If instead they fight 50 Kobolds and butcher them, they’ll think “hey, I remember when 5 of these guys nearly killed me! My character is becoming truly heroic and a mighty warrior!” They’ll then feel more prepared to take on the 5 Ogres because they know their characters may be up to the task.
    Sunday, August 31, 2008, 3:59:38 AM

    I’ve got to be honest, I was one of those who didn’t read carefully and bought into the meme for a few years. It (3rd Ed) was my first DMing experience, so I never really knew it could be different. Over time the standard 4/day at EL drove me and my players away from the system, it was just too boring and predictable.

    If only I had realized then what I know now, we all might still be playing DnD and putting money in WoTC’s pockets, instead of White Wolf’s.

    Which leads me to wonder, how many other groups ended up the same way, and how much did it cost Wizards?
    Sunday, August 31, 2008, 1:53:56 AM

    Absolutely spot on analysis, as ever. Love this.

    Too often I felt that reviewers would complain about the ELs in an adventure just to show off how clever they were – and promptly get it wrong Smile
    Saturday, August 30, 2008, 8:45:38 PM

  2. Sir Wulf says:

    I was recently contemplating the idea that player ignorance of their foes’ capabilities contributed to the idea that encounters “ought to” be balanced to the party. In classic AD&D, experienced players could easily figure out the threat level of most encounters. If you ran into 40 orcs, you could quickly figure out whether that was an encounter your group could handle. Odd or custom monsters were the exception, so players could choose when to “hold ’em and when to fold ’em”. In 3rd edition, the widespread use of classes for intelligent foes made it much tougher to “eyeball” the threat level posed by foes.

  3. delta says:

    Great, great article — thanks for pointing it out on my blog. I had Monte Cook’s “Demon God’s Fane” adventure, and was astonished that every encounter through the complex was uniformly EL 14. (If any monster wasn’t normally EL 14, it was class-boosted to just that point. Precisely as Sir Wulf says: man I hated it when a friend threw 3 kobolds in a room and class-boosted them to level 13. Ugh.) One of my greatest struggles for a decade was your item, “… more difficult and/or tedious to run the classic mega-dungeons “.

    The only point of surprise is the observation (for classic modules), “I discovered that most of the encounters in those modules would actually equate to an EL at least 2-5 levels lower than the party”. When I did the same thing (B modules, G modules, etc.), I actually found an average EL about 2-3 levels higher than the party — which I thought made sense because they expected about double the standard party size over 3E.

  4. Justin Alexander says:

    Now I’m trying to remember which modules I was actually looking at. I know it was specifically mid- and high-level stuff (so not the B series).

    And it wasn’t the G series, because the power creep of giants from 1st Edtion to 2nd Edition was maintained in 3E, which means creature-to-creature comparison is skewed. (This was a problem that cropped up when they updated the giant modules to 2E, too. AD&D1 Hill Giant = HD 8 +1. AD&D2 Hill giant = 12+1. D&D3 hill giant = 12 HD.)

    Might have been the A series. Grabbing A1, for example, the temple level encounters:

    Area 3: 8 ghouls + 2 ghasts = EL 7
    Area 6: 2 half-orcs = EL 1
    Area 6: 1 doppleganger (separate encounter) = EL 3
    Area 7: 6 orcs = EL 4
    Area 8B: 1 animated object (2 HD) = EL 2
    Area 8D: 2-8 giant worker ants = EL 3 to EL 6
    Area 8F: 2-5 ghouls = EL 3 to EL 5
    Area 9: 1 wight = EL 3
    Area 10: 2 basilisks = EL 6
    Area 12: 23 half-orcs/orcs + 3 5th level half-orcs = EL 10
    Area 14: 4 harpies = EL 7
    Area 16A: 7 half-orcs/orcs = EL 4
    Area 16B: 1 half-orc = EL 1/2
    Area 18: 6th level cleric + 4th level assassin + 3 3rd level fighters = EL 8
    Area 19: 10 stirges = EL 5 or 6

    The average level of the pregens is 5th. If I use average ELs for the encounters with an EL range, that gives me nine encounters with EL < 5, one encounter with EL 5, and five encounters with an EL 6+.

    Looking at the full level range for the adventure (4th to 7th), we can see that there are 5 encounters below the level range and only 2 above the level range. Notably, the two encounters above the level range are (a) the "boss fight" in area 18 and (b) an encounter made up almost entirely of CR 1/2 mooks.

    And, as you say, for full accuracy you'd probably want to adjust for the expected number of PCs. In this case, the module is recommended for 32 total levels (6-8 characters of 4th to 7th level). That translates to four 8th level characters, giving you: 11 encounters at APL-2 or lower; 1 encounter at APL; and 1 encounter at APL+2.

  5. Justin Alexander says:


    Wish I would have found this years ago, makes me want to write up a 3e adventure with all this in mind to see if this is what I have been missing. I have been hating 3e for years because of this sleep, fight, fight, sleep, rinse, repeat cycle. Tried Pathfinder for a while, met with the same feeling, and have all but given up on playing in exchange for yard work and home repairs. I have been reading up on a bunch of your articles, and have been loving them, thanks for taking the time to write them up.
    Thursday, April 28, 2011, 1:44:48 PM

  6. hattymchappy says:

    Wow. This article blew my mind. I didn’t start playing or know anything about D&D until a couple years after 3.5 came out. Now I play a 4e campaign and am looking to start up Pathfinder fairly soon too. I have had many problems in the past with the pace of my games when it comes to combat, especially in 4e. And it all comes down to thinking that a 4 encounter day at the party’s APL is the way to go. I am definitely going to try to build encounters the way you suggest and see how it goes from there. Thanks a lot, Justin!

  7. Greg Campbell says:

    I never religiously paid attention to CRs. When making encounters, I used things that were appropriate at the time (including things with class levels) and let my group handle ’em. They usually won after a struggle.

    I do favor lots of tough fights to train players that “life is hard so play smart or die.” (The occasional swath of low-level/CR foes is useful.)

    As for XP, I stopped giving XP and just award levels when they’re story-appropriate.

  8. Asuras says:

    On mobs… I’ve had more characters die to mobs of low level things that added up to a CR around or even slightly lower than the party’s ECL (like a bunch of CR 3 things against a 7th level party for instance) than when I throw one or two monsters against them that may add up to 3 or even 4 CR above their ECL. Mobs are great, though they can be hard to keep track of sometimes (it took some practice anyhow).

    I enjoy throwing thing of CRs all over the place against my players, probably because starting out when I drew a lot of inspiration from (though never 100% stuck to running) a few modules, you know, the old TSR ones that would do things like make a 4th level party deal with a Glabrezu, with it being pretty clearly the idea that if they fought it head on they’d all die pretty quickly.

    The other appeal of having CRs all over the place is characters really do have to think about how to deal with something, or even retreat, and if they do neither just pray they roll spectacularly well. However, having lower CR encounters that they can occasionally walk all over isn’t a bad thing either. Plus, well, the world IMO unless I’m doing strictly a board-game style dungeon crawl, naturally should have the full range of CRs to deal with, and in some cases that means you just really really want to be smart in your dealing with things, or avoid them to the best of your ability.

  9. Structured Fantasy » Blog Archive » Design Resources says:

    […] Revisiting Encounter Design […]

  10. Dan Dare says:

    Interestingly I just played a 1st level session using 5E where the players didn’t lose a single HP. Just the same they snuck into a monastery that was overrun with a war-band of 25 goblins, spoiled an alarm trap, they encounter a ghoul that was sneaking around in the crypts, and fought and defeated 6 of the goblins without raising the general alarm, found a secret treasure compartment and dealt with the ghost of the previous abbot in order to find the holy challice. Lots happened and it was very memorable, including some funny serious fumbles during the combat with the ghoul.

  11. Hedisus says:

    Minor nitpick (for me, major nitpick for people who care about proper quotation), the quotes in the header “Misreading 3rd Edition” are from pages 48 and 49 of the 3rd edition rulebook, not 100 and 101.

    Great insight into adventure design though, and I look forward to how my players will react to the new encounter mix they will face now. *Evil Grin*

  12. Justin Alexander says:

    The passages are on pgs. 100 and 101 of the 3rd Edition DMG. They’re on pgs. 48 and 49 of the 3.5 DMG.

  13. Charlie Bell says:

    Having run 3.5/PF since 3.0, I wonder if these observations still hold true. I find that the power creep inherent in late 3.5/current PF play, combined with the usual group size of 5 or 6 rather than the 4 the CR system was designed to peg, results in most CR=APL encounters being very, very easy. As in, they generally do not take more than 1 round; it is common for the monsters not to get to act before they are all killed; and it is rare that the monsters can meaningfully affect the PCs. This experience is by no means confined to highly-optimized groups, either. CR<APL encounters are hardly worth rolling the dice for.

    I have limited experience with 5E, but it seems just harder–you have to play more tactically and cannot rely on build strength to overwhelm encounters. I suspect this is an intended feature of the bounded accuracy system, but also that the bigger difference is that 5E lacks the buff stacking of 3.5/PF. I wonder if your encounter design advice is perhaps more suited to the current state of 5E than the current state of PF.

  14. ggg says:

    what are “haloscan comments”?

  15. Justin Alexander says:

    Haloscan was the commenting system on an older version of the website. When I transferred the website to WordPress, I “archived” the older comments by pasting them into WordPress comments (as seen on this post).

  16. Gamosopher says:

    Hey Justin,

    I’m curious : do you plan on doing something similar for 5e? I GMed only the beginner box (cool module), so I did not have to actually design anything, but that would be great.

Leave a Reply



Recent Posts

Recent Comments