Tuesday, July 3, 2012

[5E Musings] Monster Design

If you have been following my Playtest reports on D&D Next, then chances are good that you are aware of my dislike for the current batch of monster statblock presentation as well as design. To reiterate, this is what I said in my second session report:
I would like to take this opportunity to say that the monster statblocks are atrocious. If there's one thing I hope they take from 4E, it's the statblock design. I miss having defenses clumped together, followed by traits, and then actions. And no, you can't remove the ability scores from the shorthand versions, because ability scores are defenses in this edition! I won't go deeper into this point, though, since I feel like the playtest isn't too monster-centric yet.
With 3.5E, humanoid enemies that fulfilled the same role had differences in what their feats or NPC classes were. And in 4E, most statblocks had their own thing going. A 5E kobold can likely fill out a shorthand statblock like early 3.x modules. But so can a level 1 kobold minion from the 4E MM, if shorthands are that important. But are they? Is it really important to save space on modules as we live in a digital world?

Presentation is secondary, though, to the actual content of the monsters.

Image from WotC

Recap: Parameters of Power

Okay, admittedly, I already talked about monsters in D&D Next before. But when I did "Power=f(Level)?," the open playtest had not begun yet. We know more details now, so I believe it is time for another analysis.

But to recap, I believed, back then, that the best way to handle monsters in Next is to build up on where 4E left. That controversial edition had two parameters that handled their power level or place in the world:
  • Level: A monster's level represents his ability to interact with the player characters. An evolved form of what was the mess known as the 3.5 Challenge Ratings, it is essentially a number on the statblock that helps the DM gauge the party level that can reasonably interact with the monster.
  • Difficulty: Known as the minion-standard-elite-solo keyword axes in 4E parlance, this parameter allowed monsters of the same level to have different power levels in the game. A minion goes down in 1 hit, but are usually encountered in mobs. Standard monsters are your typical fare, elites are slightly stronger and solos should ideally be able to stand on its own.
It doesn't make much sense for a level 6 solo to face a level 12 party, for example. The solo just won't be able to hit the party, while it still has enough hit points to not go down easily. But, the same creature represented as a level 11 elite (roughly a +5/-5 scale) should be a better fit. Changing the stats to elite allows the party to feel like they've gone stronger, (wow, we used to have trouble with just one young blue dragon, but now we can take on three of them at once!) while raising its level means that the battle is still a meaningful experience. magbonch talks about the same idea in his blog.

Another way of doing things is through representing the monster throughout changing difficulties by sticking to its xp value. A level 12 solo has roughly the same amount of xp when defeated as a level 29 minion. So, reasonably, a simplified minion statblock at level 29 can represent the same level 12 solo they faced at early paragon. Upper_Krust tabulated (approximately the) same monsters in his conversion table back in his old site.

Whichever way this is done, the idea of sliding a monster up or down a level and then shifting its difficulty up or down is something that I have seen from many DMs. The only problem is that the statblocks are generally done over: Ability scores are recomputed. Powers are added if difficulty is raised and removed if difficulty is decreased. Damage is recalculated, and severity of the effects and status conditions are reconsidered. It is still a lot of hard work. If we could only avoid the hassles of recalculating scores and redoing defenses, this process would be so much easier.

Bounded Accuracy

One of the first things that many playtesters like me noticed with the first batch of playtesting was the more or less constant state of attack bonuses and defenses. Players and DMs alike immediately began to speculate that there would be no or very limited opportunities to raise these numbers. And by June 4, this Legends and Lore article confirmed the speculation: Bounded Accuracy by Rodney Thompson.

The idea seems well-received, especially after 4E made it plain that increasing numbers on both ends of the screen do not really give an actual sense of improvement for the characters. Getting rid of that eliminates the need to use monster levels to help determine if they can interact with the party; indeed, the playtest monsters did not have levels associated with them. 

However, there remains a need to measure the relative difficulty of the monster. Given a party's level, how many monsters of a certain type can the party handle?

What We Know

Monster Design has been tackled twice in L&L articles, so we have a fair idea of the ideologies behind the current design process:
  • In Monster Design Part I, Mr. Mearls talks about monster design as a four-step process. In brief: The story behind the monster is considered, analyze mechanics, add in story elements, and finally stat up the monster. 
  • In Monster Design Part II, he discusses how monsters never get out of style, and of how low powered monsters will still be relevant in higher numbers.
I applaud the implementation of the Bounded Accuracy mechanic; it is one of those things that I that I believe will help the new edition reach new heights.  However, I feel like they are milking it for all its worth, and they currently fail to see that it alone cannot save monster design.

Comparing Power Levels:

In an effort to understand the design process better, I needed to know how to compare one monster from another. But without level differences, exactly how do I do that? Well, to do that I decided to stick to xp values. After all, that's how many DMs changed the "difficulty rating" of a monster throughout the 4E level grind.
So I looked at some monsters, checked their xp values, and compared them with similarly-valued monsters as well as some who are above and below them. Here are some of my observations, both general and specific (in no particular order):
  • Ability scores have little to no relationship with xp values. They are mostly arbitrary, relative to the monster concept instead of to their power level. Of course, that also means that small, dumb kobolds (xp 75) will generally have lower ability scores than the lithe and cunning medusa (xp 450). That is, lower level monster lore will tend to mean that they have lower stats than higher level counterparts. I can dig that.
  • The same can be said of armor classes. Low ACs are just as common on either side of the xp spectrum. Now here I have some problems. It makes sense for verisimilitude to have the minotaur have low AC, but that also means that it is highly susceptible to spells like Ray of Frost, which has a strong form of control.
  • Monster HP is dependent on xp value, more or less. From 25 to 125 XP, playtest monsters will generally have a fixed xp value. From 150 to 200 XP, there aren't enough monsters to make a viable comparison. As xp's go higher, big and meaty types like the troll and minotaur tend to have less hp than complex caster types such as the medusa and the dark priest. There are exceptions, of course.
  • The stirge is simply better than the kobold, despite both being worth 75 XP. It has a slightly higher AC and hit points, which is big considering that a fighter's reaper class feature is much less likely to kill it. It has a significantly higher attack bonus (+5 versus +0). It even has a 3-saves-or-die mechanic. And what does the kobold get in return? Well, it is marginally better at damage (averaging at 2.875 versus the stirge's 2.5) and they get a ranged attack. Woop-ti-doo.
  • Skeletons and zombies differ in HP compared to other monsters at 125 XP. In the skeleton's case, a reduction in HP is probably warranted by its resistances to certain types of weapon damage. In the zombie's case, it has more hp than normal, but they suffer from reduced speeds and can't hustle. Here we experience game mechanics that compensate for story weaknesses/strengths. Okay. No complaints.
  • The kobold chieftain is a subpar choice compared to 2 elite kobolds. I used this as a comparison because two elite kobolds cost as much xp as the chief. And what does the chief get in return? Two attacks per round (with lower attack bonus and average damage compared to the elites), lower AC, and no special abilities (in the bestiary) compared to the elites who have dragonshield. He does get more HP, though.
  • Now here's something that does not make sense to me: The Dark Adept , a spellcaster at 175 XP, has more HP compared to the Berserker at 200 XP. And this is true even if you put the berserker's rage ability into account. Is this to compensate for the berserker's increased damage and immunity to fear?
  • The wight's HP is too low compared to others of the same/lower XP value. Yes, it gets enervation. But should the lethality of a monster account for a reduction in HP?
  • To reinforce the previous point: At 450 XP, the HP of caster/lethal monsters are half that of their beefy counterparts.
  • Leaders get more xp. I feel like it works on low-levels, but will tend to make less sense when you get higher up and want to have super-orcs from the Abyss that are all about as powerful as a typical orc chieftain when my players are high enough in level. This is a personal quirk, admittedly: Maybe I can just remove their more leader-y aspects to make it work.
With those observations, my personal problems with the current iteration of monsters are as follows:
  • Adding X more monsters to make things more challenging will not work for sufficiently large values of X. The widely reported cave rat problem exhibited this, but the current design direction of monsters at higher XP values also reflect it. At one point will adding more ogres be too ridiculous? They take too long to kill (imagine eight of them), but are also too easy to lock.

Screw this, there are too many of you.
  • Monsters do not necessarily have to be equal, but they should be comparable to a point that is well-represented by their XP value. The stirge-kobold comparison above is an extreme example, but the elite-chieftain kobold issue is more glaring.
  • Complexity/Lethality should not imply a low hit point value. Wait, that sounds wrong: Strong but simple monsters should not have too many hit points. 4E taught us that ramping up the hit point value does not make a solo memorable: Instead, it's the resiliency of the solo in general. It can more readily shrug off conditions (Ray of Frost!), or be affected by it less. It should have multiple opportunities to attack, and it might have a fatality/save-or-die or two. Should it have more hp than a comparable spellcaster/complex creature? Maybe, but not by too much.
  • I want to briefly address a common complaint re: monsters with spell lists. I think monster spells are okay, in moderation. The worst experience for spellslinging monsters is when their statblocks use spells from various sourcebooks, followed closely by creatures that simply have too many spells. If Fireball and Antimagic field are common spells found in a core sourcebook, I'm cool with a monster that has those prepared, with an option to prepare other similar level spells at the risk of the DM confusing himself/herself.
Revisiting The Proposal

In my first musing on Next monsters, I suggested a difficulty dial. I think it's still a good idea, and the first L&L article on monster design suggested that they plan on doing this as well. Perhaps monsters have a default difficulty setting. A few examples: 
  1. Medusas as presented by the Bestiary are difficult by default. After the statblock and the lore section, they has a few modules that shift their difficulty down. As a moderate monster, you expect to have a coven of three medusa sisters. Their save-or-suck changes into 3-saves-or-suck (like the 4E model) and their hp goes down to 50. As an easy monster, their gaze can petrify with the option for a character to unpetrify themselves through sheer constitution (con check per round). Their hp goes down to 20. Easy medusas are used when the deity Zehir is accompanied by 12 medusas.
  2. Minotaurs are standard monsters by default, and their stats are as presented in the Bestiary, except that they have 99 hit points. It is expected to be seen, in relatively large numbers, at high levels. It can also be bumped up to hard, where they are expected to be faced alone. In such cases, they have 132 hp, and their Natural Cunning is so strong that they can't be held in place by spells such as Ray of Frost. Their Charging Gore is also turned into a fatality, dealing an +10 damage. But the horns can be attacked - it has AC 18 and 30 hit points, and destroying it deals 20 damage to the minotaur. Now, easy minotaurs only have 66 hp, and their natural cunning is gone because they are expected to fight characters that are powerful enough to bypass their natural abilities.
  3. Dragonshield kobolds use the hp and ability scores of the chieftain, but use the attacks and AC of the elite kobold. They are standard monsters worth 250 xp. They can be turned to leaders by adding in a Call Reinforcements ability, and making the Dragonshield ability more selfish (grants advantage to attacks made against it 1/round).
So there, it's very rough, but the idea is to have a difficulty dial that can be adjusted depending on the style of the DM. The DM can mix and match difficulties as the players increase in levels, as what is assumed above. Alternatively a mild-handed DM might want to tone everything down to easy, or bump eveything up to hard. In either case, changing difficulties do not change the XP value of the creatures.

I don't like the current bestiary, but I do see room for improvement here. Combined with a cleaner statblock representation, I feel like a modular approach for adjusting difficulty based on play style will make for a better suite of monsters in a bestiary. Essentially, every monster has a module to adjust its difficulty, with other modules such as class or theme adjustments that are available as templates.

What do you think?